Silicon ChipFebruary 1994 - Silicon Chip Online SILICON CHIP
  1. Outer Front Cover
  2. Contents
  3. Publisher's Letter: Energy consumption taking the long view
  4. Feature: Airbags: More Than Just Bags Of Wind by Julian Edgar
  5. Book Store
  6. Feature: Data On The ISD2590P Voice Recorder IC by Darren Yates
  7. Project: Build A 90-Second Message Recorder by Darren Yates
  8. Feature: Instrumentation Programming The Graphical Way by Jack Barber
  9. Project: Compact & Efficient 12-240VAC 200W Inverter by John Clarke
  10. Feature: Electronic Engine Management; Pt.5 by Julian Edgar
  11. Project: A Single Chip Audio Amplifier by Darren Yates
  12. Serviceman's Log: If only the fault would show by The TV Serviceman
  13. Feature: Amateur Radio by Garry Cratt, VK2YBX
  14. Project: Build A Novel LED Torch by John Clarke
  15. Order Form
  16. Project: 40V 3A Variable Power Supply; Pt.2 by John Clarke
  17. Feature: Computer Bits by Darren Yates
  18. Vintage Radio: Building a simple 1-valve receiver by John Hill
  19. Product Showcase
  20. Back Issues
  21. Market Centre
  22. Advertising Index
  23. Outer Back Cover

This is only a preview of the February 1994 issue of Silicon Chip.

You can view 30 of the 96 pages in the full issue, including the advertisments.

For full access, purchase the issue for $10.00 or subscribe for access to the latest issues.

Items relevant to "Build A 90-Second Message Recorder":
  • 90 Second Message Recorder PCB pattern (PDF download) [01202941] (Free)
Items relevant to "Compact & Efficient 12-240VAC 200W Inverter":
  • 200W 12V DC to 240VAC Inverter PCB pattern (PDF download) [11309931] (Free)
Articles in this series:
  • Electronic Engine Management; Pt.1 (October 1993)
  • Electronic Engine Management; Pt.1 (October 1993)
  • Electronic Engine Management; Pt.2 (November 1993)
  • Electronic Engine Management; Pt.2 (November 1993)
  • Electronic Engine Management; Pt.3 (December 1993)
  • Electronic Engine Management; Pt.3 (December 1993)
  • Electronic Engine Management; Pt.4 (January 1994)
  • Electronic Engine Management; Pt.4 (January 1994)
  • Electronic Engine Management; Pt.5 (February 1994)
  • Electronic Engine Management; Pt.5 (February 1994)
  • Electronic Engine Management; Pt.6 (March 1994)
  • Electronic Engine Management; Pt.6 (March 1994)
  • Electronic Engine Management; Pt.7 (April 1994)
  • Electronic Engine Management; Pt.7 (April 1994)
  • Electronic Engine Management; Pt.8 (May 1994)
  • Electronic Engine Management; Pt.8 (May 1994)
  • Electronic Engine Management; Pt.9 (June 1994)
  • Electronic Engine Management; Pt.9 (June 1994)
  • Electronic Engine Management; Pt.10 (July 1994)
  • Electronic Engine Management; Pt.10 (July 1994)
  • Electronic Engine Management; Pt.11 (August 1994)
  • Electronic Engine Management; Pt.11 (August 1994)
  • Electronic Engine Management; Pt.12 (September 1994)
  • Electronic Engine Management; Pt.12 (September 1994)
  • Electronic Engine Management; Pt.13 (October 1994)
  • Electronic Engine Management; Pt.13 (October 1994)
Items relevant to "A Single Chip Audio Amplifier":
  • Champ: Single Chip Audio Amplifier PCB [01102941] (AUD $5.00)
  • The Champ PCB pattern (PDF download) [01102941] (Free)
Articles in this series:
  • Amateur Radio (November 1987)
  • Amateur Radio (November 1987)
  • Amateur Radio (December 1987)
  • Amateur Radio (December 1987)
  • Amateur Radio (February 1988)
  • Amateur Radio (February 1988)
  • Amateur Radio (March 1988)
  • Amateur Radio (March 1988)
  • Amateur Radio (April 1988)
  • Amateur Radio (April 1988)
  • Amateur Radio (May 1988)
  • Amateur Radio (May 1988)
  • Amateur Radio (June 1988)
  • Amateur Radio (June 1988)
  • Amateur Radio (July 1988)
  • Amateur Radio (July 1988)
  • Amateur Radio (August 1988)
  • Amateur Radio (August 1988)
  • Amateur Radio (September 1988)
  • Amateur Radio (September 1988)
  • Amateur Radio (October 1988)
  • Amateur Radio (October 1988)
  • Amateur Radio (November 1988)
  • Amateur Radio (November 1988)
  • Amateur Radio (December 1988)
  • Amateur Radio (December 1988)
  • Amateur Radio (January 1989)
  • Amateur Radio (January 1989)
  • Amateur Radio (April 1989)
  • Amateur Radio (April 1989)
  • Amateur Radio (May 1989)
  • Amateur Radio (May 1989)
  • Amateur Radio (June 1989)
  • Amateur Radio (June 1989)
  • Amateur Radio (July 1989)
  • Amateur Radio (July 1989)
  • Amateur Radio (August 1989)
  • Amateur Radio (August 1989)
  • Amateur Radio (September 1989)
  • Amateur Radio (September 1989)
  • Amateur Radio (October 1989)
  • Amateur Radio (October 1989)
  • Amateur Radio (November 1989)
  • Amateur Radio (November 1989)
  • Amateur Radio (December 1989)
  • Amateur Radio (December 1989)
  • Amateur Radio (February 1990)
  • Amateur Radio (February 1990)
  • Amateur Radio (March 1990)
  • Amateur Radio (March 1990)
  • Amateur Radio (April 1990)
  • Amateur Radio (April 1990)
  • Amateur Radio (May 1990)
  • Amateur Radio (May 1990)
  • Amateur Radio (June 1990)
  • Amateur Radio (June 1990)
  • Amateur Radio (July 1990)
  • Amateur Radio (July 1990)
  • The "Tube" vs. The Microchip (August 1990)
  • The "Tube" vs. The Microchip (August 1990)
  • Amateur Radio (September 1990)
  • Amateur Radio (September 1990)
  • Amateur Radio (October 1990)
  • Amateur Radio (October 1990)
  • Amateur Radio (November 1990)
  • Amateur Radio (November 1990)
  • Amateur Radio (December 1990)
  • Amateur Radio (December 1990)
  • Amateur Radio (January 1991)
  • Amateur Radio (January 1991)
  • Amateur Radio (February 1991)
  • Amateur Radio (February 1991)
  • Amateur Radio (March 1991)
  • Amateur Radio (March 1991)
  • Amateur Radio (April 1991)
  • Amateur Radio (April 1991)
  • Amateur Radio (May 1991)
  • Amateur Radio (May 1991)
  • Amateur Radio (June 1991)
  • Amateur Radio (June 1991)
  • Amateur Radio (July 1991)
  • Amateur Radio (July 1991)
  • Amateur Radio (August 1991)
  • Amateur Radio (August 1991)
  • Amateur Radio (September 1991)
  • Amateur Radio (September 1991)
  • Amateur Radio (October 1991)
  • Amateur Radio (October 1991)
  • Amateur Radio (November 1991)
  • Amateur Radio (November 1991)
  • Amateur Radio (January 1992)
  • Amateur Radio (January 1992)
  • Amateur Radio (February 1992)
  • Amateur Radio (February 1992)
  • Amateur Radio (March 1992)
  • Amateur Radio (March 1992)
  • Amateur Radio (July 1992)
  • Amateur Radio (July 1992)
  • Amateur Radio (August 1992)
  • Amateur Radio (August 1992)
  • Amateur Radio (September 1992)
  • Amateur Radio (September 1992)
  • Amateur Radio (October 1992)
  • Amateur Radio (October 1992)
  • Amateur Radio (November 1992)
  • Amateur Radio (November 1992)
  • Amateur Radio (January 1993)
  • Amateur Radio (January 1993)
  • Amateur Radio (March 1993)
  • Amateur Radio (March 1993)
  • Amateur Radio (May 1993)
  • Amateur Radio (May 1993)
  • Amateur Radio (June 1993)
  • Amateur Radio (June 1993)
  • Amateur Radio (July 1993)
  • Amateur Radio (July 1993)
  • Amateur Radio (August 1993)
  • Amateur Radio (August 1993)
  • Amateur Radio (September 1993)
  • Amateur Radio (September 1993)
  • Amateur Radio (October 1993)
  • Amateur Radio (October 1993)
  • Amateur Radio (December 1993)
  • Amateur Radio (December 1993)
  • Amateur Radio (February 1994)
  • Amateur Radio (February 1994)
  • Amateur Radio (March 1994)
  • Amateur Radio (March 1994)
  • Amateur Radio (May 1994)
  • Amateur Radio (May 1994)
  • Amateur Radio (June 1994)
  • Amateur Radio (June 1994)
  • Amateur Radio (September 1994)
  • Amateur Radio (September 1994)
  • Amateur Radio (December 1994)
  • Amateur Radio (December 1994)
  • Amateur Radio (January 1995)
  • Amateur Radio (January 1995)
  • CB Radio Can Now Transmit Data (March 2001)
  • CB Radio Can Now Transmit Data (March 2001)
  • What's On Offer In "Walkie Talkies" (March 2001)
  • What's On Offer In "Walkie Talkies" (March 2001)
  • Stressless Wireless (October 2004)
  • Stressless Wireless (October 2004)
  • WiNRADiO: Marrying A Radio Receiver To A PC (January 2007)
  • WiNRADiO: Marrying A Radio Receiver To A PC (January 2007)
  • “Degen” Synthesised HF Communications Receiver (January 2007)
  • “Degen” Synthesised HF Communications Receiver (January 2007)
  • PICAXE-08M 433MHz Data Transceiver (October 2008)
  • PICAXE-08M 433MHz Data Transceiver (October 2008)
  • Half-Duplex With HopeRF’s HM-TR UHF Transceivers (April 2009)
  • Half-Duplex With HopeRF’s HM-TR UHF Transceivers (April 2009)
  • Dorji 433MHz Wireless Data Modules (January 2012)
  • Dorji 433MHz Wireless Data Modules (January 2012)
Items relevant to "Build A Novel LED Torch":
  • Simple LED Torch PCB pattern (PDF download) [08302941] (Free)
Items relevant to "40V 3A Variable Power Supply; Pt.2":
  • 40V 3A Variable Power Supply PCB pattern (PDF download) [04202941] (Free)
Articles in this series:
  • 40V 3A Variable Power Supply; Pt.1 (January 1994)
  • 40V 3A Variable Power Supply; Pt.1 (January 1994)
  • 40V 3A Variable Power Supply; Pt.2 (February 1994)
  • 40V 3A Variable Power Supply; Pt.2 (February 1994)
Items relevant to "Computer Bits":
  • Games Card Breakout PCB [07103941] (PCB Pattern, Free)
  • DOS software for Computer Bits, November 1993 & January/February 1994 (Free)
Articles in this series:
  • Experiments For Your Games Card (January 1992)
  • Experiments For Your Games Card (January 1992)
  • Computer Bits (November 1993)
  • Computer Bits (November 1993)
  • Computer Bits (January 1994)
  • Computer Bits (January 1994)
  • Computer Bits (February 1994)
  • Computer Bits (February 1994)
  • Computer Bits (April 1994)
  • Computer Bits (April 1994)
Vol.7, No.2; February 1994 FEATURES FEATURES   4 Airbags: More Than Just Bags Of Wind by Julian Edgar AIRBAGS ARE MORE than just bags of wind that operate on a hit or miss basis. Our article on page 4 explains how they work. Find out how they work 10 Data On The ISD2590P Voice Recorder IC by Darren Yates New chip has up to 90 seconds playback time 22 Instrumentation Programming The Graphical Way by Jack Barber Graphical programming using LabVIEW 42 Electronic Engine Management, Pt.5 by Julian Edgar The oxygen sensor – how it works PROJECTS PROJECTS TO TO BUILD BUILD 16 Build A 90-Second Message Recorder by Darren Yates THIS 90-SECOND MESSAGE recorder is based on a special IC & features a pause button, zero-power memory retention & battery operation. Construction starts on page 16. Features a pause button & zero-power memory storage 26 Compact & Efficient 12-240VAC 200W Inverter by John Clarke New switching design has low stand-by current 46 A Single Chip Audio Amplifier by Darren Yates Compact design delivers 0.5W into eight ohms 58 Build A Novel LED Torch by John Clarke Uses a high-brightness amber LED for low battery drain 66 40V 3A Variable Power Supply, Pt.2 by John Clarke Construction, test & adjustment SPECIAL SPECIAL COLUMNS COLUMNS FEATURING A HIGHLY efficient switching circuit, this compact 12-240VAC inverter can drive most mains appliances, including TVs, computers, power tools & fluorescent lights. See page 26. 50 Serviceman’s Log by the TV Serviceman If only the fault would show 56 Amateur Radio by Garry Cratt, VK2YBX Convert an inexpensive walkie-talkie to the 6-metre band 79 Computer Bits by Darren Yates Experiments for your games card, Pt.4 82 Vintage Radio by John Hill Building a simple 1-valve receiver DEPARTMENTS DEPARTMENTS   2 20 65 87 Publisher’s Letter Circuit Notebook Order Form Product Showcase 90 92 95 96 Back Issues Ask Silicon Chip Market Centre Advertising Index THE OXYGEN SENSOR is a vital component in your car’s engine management system. Find out what it does & how it works by turning to page 42. Cover design: Marque Crozman February 1994  1 Publisher & Editor-in-Chief Leo Simpson, B.Bus. Editor Greg Swain, B.Sc.(Hons.) Technical Staff John Clarke, B.E.(Elec.) Robert Flynn Darren Yates, B.Sc. Reader Services Ann Jenkinson Sharon Macdonald Marketing Manager Sharon Lightner Phone (02) 979 5644 Mobile phone (018) 28 5532 Regular Contributors Brendan Akhurst Garry Cratt, VK2YBX Marque Crozman, VK2ZLZ John Hill Jim Lawler, MTETIA Bryan Maher, M.E., B.Sc. Philip Watson, MIREE, VK2ZPW Jim Yalden, VK2YGY Bob Young Photography Stuart Bryce SILICON CHIP is published 12 times a year by Silicon Chip Publications Pty Ltd. A.C.N. 003 205 490. All material copyright ©. No part of this publication may be reproduced without the written consent of the publisher. Printing: Macquarie Print, Dubbo, NSW. Distribution: Network Distribution Company. Subscription rates: $49 per year in Australia. For overseas rates, see the subscription page in this issue. Editorial & advertising offices: Unit 34, 1-3 Jubilee Avenue, Warrie­ wood, NSW 2102. Postal address: PO Box 139, Collaroy Beach, NSW 2097. Phone (02) 979 5644. Fax (02) 979 6503. ISSN 1030-2662 PUBLISHER'S LETTER Energy consumption: taking the long view So the Federal Opposition has decided to scrap its policy of supporting nuclear energy plants and has admitted that its former position had proved impractical. After at least a decade of espousing nuclear energy they have finally seen the light. Now what they need is an enlightened energy policy which attempts to change community attitudes rather than reacting to them. An enlightened energy policy from the Federal Opposition would seek to encourage conservation in all aspects of energy use, particularly with respect to electricity. Almost everywhere you look, whether in business or domestic consumption, there is huge energy waste. In commercial buildings, retail stores and industry, there is plenty of scope for reducing this waste. Most commercial and retail establishments, for example, have excessive and often ineffective lighting, and airconditioned buildings with no provision for reducing solar heat build-up. And even during the recent recession, these were fully lit up at all hours of the night. Domestic use of electricity is also extremely wasteful. Many people have excessive lighting inside and outside their homes; they often run two or three refrigerators when one would be adequate and when you add in the energy consumption associated with a swimming pool, their energy consumption over a year must run into many megawatt-hours. Now while you might think that more energy use equates to more employment, in the long run such waste of energy cannot do the economy, or the environment, any good. Excessive energy use means that more expensive power stations have to be constructed and paid for, more coal has to be mined (usually open cut), and then more work has to be done to restore the landscape after mining. Anyone who has seen the huge ash dumps associated with our power stations cannot fail to be awed by the extent of this problem. Clearly, any policy to encourage efficient use of energy will not be produced overnight and aspects of it are likely to be unpopular – higher tariffs or a carbon tax, for example. But as with other energy sources which are clearly running low, such as petroleum, Australia cannot afford to be profligate forever. Having said that, the Federal Opposition and indeed, the Government, should take a long hard look at their energy policy and, while they’re at it, make sure they encourage the develop­ment of solar energy – it’s our only inexhaustible energy re­source. Leo Simpson Please note our new address: Unit 34, 1-3 Jubilee Avenue, Warriewood, NSW 2102. Our telephone/fax numbers and postal address remain the same. WARNING! SILICON CHIP magazine regularly describes projects which employ a mains power supply or produce high voltage. All such projects should be considered dangerous or even lethal if not used safely. Readers are warned that high voltage wiring should be carried out according to the instructions in the articles. When working on these projects use extreme care to ensure that you do not accidentally come into contact with mains AC voltages or high voltage DC. If you are not confident about working with projects employing mains voltages or other high voltages, you are advised not to attempt work on them. Silicon Chip Publications Pty Ltd disclaims any liability for damages should anyone be killed or injured while working on a project or circuit described in any issue of SILICON CHIP magazine. Devices or circuits described in SILICON CHIP may be covered by patents. SILICON CHIP disclaims any liability for the infringement of such patents by the manufacturing or selling of any such equipment. SILICON CHIP also disclaims any liability for projects which are used in such a way as to infringe relevant government regulations and by-laws. Advertisers are warned that they are responsible for the content of all advertisements and that they must conform to the Trade Practices Act 1974 or as subsequently amended and to any governmental regulations which are applicable. 2  Silicon Chip High Purchase Costs Taking a “Bite” Out of Your Budget? NOT AT MACSERVICE. WE HELP YOU STRIKE BACK BY OFFERING THE LOWEST PRICES AND GOOD OLD FASHIONED SERVICE - Just look at these SPECIALS BALL EFRATOM M100 Rubidium Frequency • Factory cal. certs. • Perfect for ISO    accreditation • GPS applications • Ruggedised military    design TEKTRONIX 5440 Oscilloscope • DC to 60MHz • 1mV - 100V/div (x 10) • Dual Trace • Dual Timebase • Large Screen TEKTRONIX 7603 Oscilloscope • Mil spec AN/USM 281-C • Triggers to 100MHz • Dual Trace • Dual Timebase • Large Screen SUPER SALE $850 GREAT VALUE $2950 (new) Video Dist Amp & Cable Equaliser   $100 ADVANCE PP7 30V3A DC Power Supply   $150 AVO MK.IV Avometer With Cal.   $275 BPL CB154/4 Electrolytic Cap Bridge   $450 B&K 1466A 10MHz Oscilloscope   $275 EH 129 Pulse Generator   $90 ELGENCO 603A White Noise Gen 5MHz   $200 ENI 503L RF Power Amp 40dB 510MHz $1025 FLUKE 102 VAW Cal Meter    $75 FLUKE 9010A Logic System Troubleshooter $1000 GR 1608 LCR Meter – Lab Standard $1500 HP 211B 10MHz Square Wave Generator   $275 HP 302A Audio Selective Level Meter   $145 HP 400L True RMS Voltmeter   $170 HP 410B Vacuum Tube Voltmeter   $130 HP 432A 10GHz Power Meter (c/w sensor)   $875 SUPER DEAL $950 HP HP HP HP HP HP HP HP I/S Elect. MARCONI MARCONI MARCONI MARCONI MARCONI MARCONI MARCONI HEWLETT PACKARD HEWLETT PACKARD 200CD Audio Oscillator 410C Multimeter • 5Hz to 600kHz • 100Hz to 700MHz • 5 Ranges • AC/DC Volts • 10V Out • DC Amps • Balanced Output • 10 ohms to 10M ohms • Complete with probes BARGAIN $265 RIDICULOUS $79 467A Power Amp   $175 536A Frequency Meter   $75 721A 30V 0.3A Power Supply    $60 1610B Logic Analyser   $450 1980 100MHz Storage Oscilloscope $1650 3400A True rms voltmeter   $425 6226A Power Supply 40V 1.5A   $200 54111D Ultimate Digital Storage Scope $19000 845 Prog Function Generator   $800 TF893A Power Meter   $150 TF1020A RF Power Meter 75Ω 100W    $75 TF1020A-1 RF Power Meter 50Ω 100W   $150 TF1245/46/47 Q Meter 40KHz-300MHz   $600 TF2167 RF Amplifier 47dB gain   $600 TF2300 FM/AM Mod Meter   $300 TF2300A FM/AM Mod Meter   $495 MARCONI TF2300B MARCONI TF2303 MARCONI TF2700 MARCONI TF2701 MARCONI TF2914 PACIFIC PM1017 RACAL 9500 SHALLTRONIX 10K SIEMENS G2212 SIEMENS P2005 SOLA Series 200 Spectral Dyn. SD112-1 Systron Don. 1037 Telequipment CT71 TRIMAX G1B VARIAC Mod Meter 1200MHz $1100 AM/FM Mod Meter   $550 LCR Bridge   $325 Universal Bridge in circuit   $700 Insertion Signal Analyser   $150 Log Freq-Voltage Converter   $150 100MHz GPIB Counter   $350 Decade Box   $150 1.6/18.6MHz Generator   $250 Controllable Phase Meter   $200 750VA Line Stabiliser   $180 Voltmeter Freq-Log Conv 2ch   $150 500MHz Counter   $350 Curve Tracer   $900 Ionisation Tester 10kV   $260 0/280V <at> 15A   $260 NEW METROLOGY INSTRUMENTS AT FANTASTIC PRICES!!! M36 $55 VCE 150 $120 CM 25 $45 SEPTEMBER SPECIAL TEKTRONIX 465M 100MHz Oscilloscope VCE-150 VCE-200 VCD-150 DI-10 DI-1 TDI-0.8 CM-25 CM-50 150mm/6" Electronic Digital Vernier in box $120 200mm/8" Electronic Digital Vernier in box $180 150mm x 0.02 Dial Vernier Caliper   $75 10 x 0.01mm Dial Indicator   $45 1" x 0.001" Dial Indicator   $45 0-0.8 x 0.01mm Test Dial Indicator   $95 0-25mm x 0.01mm Outside Micrometer   $45 25-50mm x 0.01mm Outside Micrometer   $55 The Name That Means Quality CM-75 50-75mm x 0.01mm Outside Micrometer   $65 CM-01 0-1" x0.001" Outside Micrometer   $45 MB-6 CZ-6C Magnetic Base Stand   $55 VC-150 Dual Scale Vernier Caliper 150 x 0.02mm/6" x 0.001"   $35 VC-200* Dual Scale Vernier Caloper 200 x 0.02mm/8" x 0.001"   $45 VC-600* Dual Scale Vernier Caliper 600 x 0.02mm/24" x 0.001" $250 HI-600 600mm/24" x 0.02mm Height Gauge $550 *WITH FINE ADJUSTMENT Affordable Laboratory Instruments SSI-2360 60MHz Dual Trace Dual Timebase Oscilloscope BRA BRAN D EQUIP NEW MENT ND EQUIP NEW MENT Bandwidth DC to 100MHz; Rise time <=3.5ns; Deflection factor 5mV/div to 5V/ div in 10 steps; DC accuracy ±2%; 2-channel display mode; Horizontal deflection - main & delayed timebases; A - 0.5s/div to 0.05µs/div in 22 steps; B - 50ms/div to 0.05µs/div in 19 steps; Trigger - main/delay sweep; Coupling AC, DC, LF Rejection, HF Rejection TOP VALUE $1150 • • • • • • 60MHz dual trace, dual trigger Vertical sensitivity 1mV/div. Maximum sweep rate 5ns/div. Built-in component tester With delay sweep, single sweep Two high quality probes $1050 + Tax PS303D Dual Output Supply • 0 to 30V and 0 to 3 amps • Four output meters • Independent or Tracking modes • Low ripple output $385 + Tax PS303 Single Output Supply PS305D Dual Output Supply PS305 Single Output Supply • 0 to 30V and 0 to 5 amps $430 + Tax • 0 to 30V and 0 to 3 amps • Two output meters • Constant current/voltage • Low ripple output $225 + Tax • 0 to 30V and 0 to 5 amps $260 + Tax IF IT’S NOT HERE WE CAN GET IT... CALL US FIRST OR CALL US LAST... BUT DON’T FORGET TO CALL US! MACSERVICE Australia’s Largest Remarketer of Test & Measurement Equipment 26 Fulton Street, Oakleigh Sth, Vic., 3167   Tel: (03) 562 9500 Fax: (03) 562 9615 **Illustrations are representative only (1) (2) Airbags: more than just a bag of wind Although widely used in the US, Japan & Europe, vehicle airbags have only recently become popular in Australia. Here’s a rundown on how they work. By JULIAN EDGAR For those who have not seen the publicity surrounding their Australian introduction, the airbag (or Supplementary Restraint System – SRS) is a cushion which inflates out of the centre of the steering wheel (or dashboard) in the event of a front-end accident. The idea is to cushion the impact and prevent (or at the very least significantly reduce) injuries to the head and chest area of the victim. A typical airbag system is that introduced by Holden in its VR Commodore. Fig.1 shows the layout of the device. In the Commo­dore, a single airbag is fitted on the driver’s side while in some other cars, a passenger-side airbag is also fitted. Side impact airbags are currently being trialled by some manufactur­ers. Bag inflation COVER AIRBAG INFLATOR STEERING WHEEL CLOCK SPRING COIL CRASH SENSOR FIG.1: THE MAJOR components in the VR Commodore airbag system. The airbag inflates out of the centre of the steering wheel. (Courtesy General Motors Holdens). 4  Silicon Chip The Holden airbag is constructed of silicone-coated nylon. It has a volume of 65 litres and is 700mm in diameter. When trig­gered, it inflates in just 30 milliseconds and the bag then deflates THE INTERNALS OF the current Bosch airbag trigger, as used in the Holden VR Commodore. The circuit board on the right has been folded out for this photo; normally it is stacked above the other board. FACING PAGE: (1) The Holden Commodore airbag (shown here deployed in a promotion­al photo for the Toyota Lexcen) has a volume of 65 litres & takes just 30 milliseconds to inflate. (2) Toyota’s new Tarago Ultima & GLX vehicles have a driver’s side airbag as standard equipment. This is what it looks like when fully inflated. in about 100 milliseconds as the driver impacts it (by way of comparison, a blink of an eye also typically takes about 100 milliseconds). The deflation speed is controlled by providing two 45mm vents in the bag, while the initial expansion rate of the bag is controlled by the use of two internal tethers, which stop the bag from head-butting you before you hit it! A sodium azide gas generator is what causes the airbag to inflate so rapidly. This airbag inflator – located within the hub of the steering wheel –is triggered by a crash sensor via a “clock spring coil”, a device that does away with the need for slip rings. This is used because the necessary reaction time of the airbag is so short that sliprings (like those used for the horn, for example) are not reliable enough – one contact might be momentarily lifted at the time of impact and so the airbag would not trigger at precisely the required moment. Fig.2 shows the relationship between vehicle deformation, driver movement and airbag inflation in the Commodore. Triggering the action of the airbag is an electronic sen­sor. Just consider for a moment the magnitude of the task facing the designers of this sensor. To begin with, the “ideal” sensor must discriminate between a crash and a parking bump or driving over a gutter. FIG.2 (BELOW): the sequence of events during a crash. (Courtesy General Motors Hold­ens). Impact The crash begins when the front of the bumper contacts the impacting object. In the next 15ms the crash sensor determines the severity of the collision & decides whether to deploy the airbag. Burst out The airbag housed in the centre of the steering wheel splits its covering pad in predetermined places & begins to inflate rapidly. Inflation The airbag is now fully inflated as the driver begins to move forward. The seatbelt progressively restricts the driver’s forward movement. Contact The driver’s head & chest contact the airbag & it immediately begins to deflate. The large area of the bag evenly distributes head & chest loads thereby significantly reducing the risk of severe injury. Support The driver sinks deep into the continually deflating airbag & upon reaching the limit of forward movement, begins to rebound. Rebound The driver continues to travel rearwards until making contact with the seat back & head restraint. February 1994  5 THE TWO PIEZO accelerometers are contained within the metal housing. Note the rigid attachment of the accelerometer module to the cast aluminium chassis. It must also be totally reliable, totally immune to false triggering, and it must be capable of firing the airbag even if the normal battery supply has been lost during the im­pact. Finally, it is also preferable if it can detect any inter­ nal faults in the system, either within the sensor itself or in the airbag inflator. so-called “ball in a tube” sensor. This elec­ tro-mechanical sensor consists of a glass tube, with a steel ball held in place at one end by a magnet. Two electrical contacts are located at the other end and the tube is filled with a gas damp­ing medium. If a crash occurs, the rapid deceleration of the ball over­ comes the attraction of the magnet. The ball thus rockets down to the other end of the tube and shorts the electrical contacts, thereby causing the airbag to inflate. This crude sensor is now rarely used. To be effective, it needed to be Old-style triggers A variety of sensors has been used over the years – none of which had the capabilities of the “ideal” sensor described above. The simplest is the VIGN LAMP TEST ACCELEROMETER 1 ACCELEROMETER 2 P SQUIBS CPU TEST VOLTAGE REGULATOR N WATCH DOG V VIGN DELAY ENERGY ANALOG INHIBIT EXTERNAL SWITCHES INHIBIT FIG.3: BLOCK DIAGRAM of the Bosch airbag trigger sensor. It uses a micro­ controller to monitor the outputs from two accelerometers & has various other circuits to prevent false triggering. The airbag is triggered by simultaneously switching on two output transistors. 6  Silicon Chip located towards the front of the vehicle, otherwise the cushioning affect provided by the vehicle’s body as it crushed delayed the triggering action. However, a frontal loca­ tion caused problems in terms of the vulnerable wiring needed to connect it to the airbag. Tuning the electro-mechanical sensor was also difficult. Electronic crash sensors were then brought into use. One Bosch sensor used a strain gauge attached to a pendulum which was suspended in a damping medium. A calculated acceleration of 4G (about the same as occurs during a frontal impact at 15km/h) was required for the sensor to fire the airbag. However, the unreli­ ability of this type of sensor meant that a device such as a mercury switch was usually placed in series with it to prevent the bag from activating under normal operating conditions. Generally, in this type of system, the mechanically inte­grating sensors were placed within the crush zone and worked in conjunction with a centrally-placed electronic sensor. The latest sensor The Bosch electronic sensor currently in use is much more sophisticated than either of the above sensors. It incorporates all of the characteristics of the “ideal” sensor mentioned above and also includes crash event data-logging and a serial data link. It is also fully programmable, allowing it to be calibrated for different vehicles. Fig.3 shows a block diagram of the sensor. As shown, the sensor uses two accelerometers which are based on piezoelectric transducers. The sensing element consists of two reverse polarized piezo oxide bars with two electrodes each. These are cemented together and form a bimorph element. During deceleration, one bar is compressed and the other stretched. Because the two bars are reverse polarised, the sum of their individual voltages appears between the two outer elec­trodes; ie, the signal is effectively doubled, thus giving good sensitivity. A low-pass filter with a cut-off frequency of 300Hz is used between the sensor and its amplifier. This filters out the 10kHz resonance peak of the sensor and avoids signal distortion when the output signal is sampled by the mi- FIG.4: A TYPICAL sensor output during a crash. The microcontrol­ler’s algorithm is used to derive the core deceleration from the high frequency variations. Time T0 is the start of the crash, T1 is the beginning of the airbag inflation, and T2 is when the airbag is deflating under the impact of the occupant. Heavy braking (just prior to wheel lock-up) in a road car develops only about 0.9G deceleration crocontroller. The sensor’s amplifier is built to work within the somewhat mind-boggling range of ±35G! A crash is detected by using a microcontroller to sample the sensor output, perform an analog/digital conversion, and then integrate this value with respect to time. If the derived value exceeds a certain threshold, the airbag will be fired. However, this integration is not sufficient to discriminate between all crashes. Oblique impacts, offset crash­es, centre-pole crashes and slow frontal barrier crashes all cause problems with this approach. Further data processing is therefore superimposed on the straight integration to improve crash discrimination. Two separate channels are used, with each accelerometer monitored. For the bag to be fired, an “interval watchdog” must receive triggering pulses from each of the two signal processing programs. If one program is not working properly, then the watch­ dog detects the missing triggering edge and inhibits the output stages. The other important role which the FIRST CRASH TESTS WITH TARGET VEHICLE ANALYSIS OF CRASH DATA ADJUSTMENT OF DEPLOYMENT ALGORITHM TO TARGET VEHICLE COMPUTER SIMULATION OF DEPLOYMENT REQUIRED FIRING TIMES ACHIEVED? N Y PROGRAM TEST - ECU CRASH TEST WITH TARGET VEHICLE. FINAL VERIFICATION FIG.5: TYPICAL airbag sensor calibration flow chart. Crash testing plays an important role. sensor must play is in predicting the deceleration that the car will experience during the inflation time of the airbag. If the airbag inflates too late, then the crash victim will already be in contact with the bag as it expands. This could lead to a situation where the victim could actually suffer an increase in acceleration – in the opposite direction! During a crash, there are high frequency variations in the deceleration superimposed on a ramping curve. Tests with dummies have shown that these high frequency variations have little effect on the dummy’s “health” – it’s the core signal of low frequency deceleration which is vital. The algorithm must there­fore smooth the accelerometer’s output to obtain the core signal and then predict the magnitude of this core signal during the period that the airbag is inflating. Fig.4 shows the modulated and core deceleration signals derived from the accelerometers. Output stage The sensor’s output stages to the airbag inflator – or “squib” – are shown in Fig.3 and use two power transistors to fire the airbag. At the start of a crash, the microcontroller sends a trigger enable signal and – after a small delay – the output stages are enabled. If February 1994  7 program and some of these would have provided data to calibrate the airbag sensor (among other things). Testing is also carried out to ensure that the airbag can not be triggered by a hammer-blow or by driving along a rough road. Any unexpected inflation of the airbag could cause the driver to crash. Fault codes & data logging VOLKSWAGEN BARRIER testing of an airbag. Note the seatbelt stretch. Bosch state that in any impact over about 40km/h, the driver will impact the steering wheel, even when wearing a seat­belt. the crash is of sufficient magni­tude, both power transistors are switched on to close the firing loop and inflate the airbag. The firing squib is constantly monitored for inappropriate electrical conditions (like squib resistance change) and the power transistors are tested each time the car is started by sequent­ial­ly switching them on for a short time. Power reserve If the main power supply to the sensor module is disrupted during a crash, an on-board “energy reserve 8  Silicon Chip capacitor” is used as the power source instead. This power source is also constantly moni­ tored for fault conditions. Sensor calibration Calibrating the sensor to suit a specific vehicle is vital. Actual crash testing of a car into a barrier is expensive and so computer modelling is extensively used to reduce the number of test crashes required. Fig.5 shows a typical sensor calibration flow chart. Holden crashed 45 cars into a concrete barrier as part of the VR Commodore development If a fault is detected by the module, either in the sensor itself or in the airbag inflator, a warning light is illuminated on the dashboard. A corresponding fault code is also stored in non-volatile memory. The non-volatile memory is also used to store information generated during the crash itself. Stored within the EEPROM are samples of the deceleration signals encountered during the crash, the time interval between the start of the crash and the deploy­ment of the bag, any errors detected before and during the crash, and the elapsed time since the warning light had last been switched on. A study of some of the G forces recorded in EEPROMs during actual crashes might reveal some sobering statistics and could help improve SC vehicle design. Acknowledgements Thanks to Robert Bosch Australia and General Motors Hold­ens for supplying the information used in compiling this article. SILICON CHIP BOOK SHOP Newnes Guide to Satellite TV 336 pages, in paperback at $49.95. Installation, Recept­ion & Repair. By Derek J. Stephen­son. First published 1991, reprinted 1994 (3rd edition). This is a practical guide on the installation and servicing of satellite television equipment. The coverage of the subject is extensive, without excessive theory or mathematics. 371 pages, in hard cover at $55.95. Servicing Personal Computers By Michael Tooley. First pub­ lished 1985. 4th edition 1994. Computers are prone to failure from a number of common causes & some that are not so common. This book sets out the principles & practice of computer servicing (including disc drives, printers & monitors), describes some of the latest software diagnostic routines & includes program listings. 387 pages in hard cover at $59.95. The Art of Linear Electronics By John Linsley Hood. Pub­lished 1993. This is a practical handbook from one of the world’s most prolific audio designers, with many of his designs having been published in English technical magazines over the years. A great many practical circuits are featured – a must for anyone inter­ested in audio design. Optoelectronics: An Introduction By J. C. A. Chaimowicz. First published 1989, reprinted 1992. This particular field is about to explode and it is most important for engineers and technicians to bring themselves up to date. The subject is comprehensively covered, starting with optics and then moving into all aspects of fibre optic communications. 361 pages, in paperback at $55.95. Digital Audio & Compact Disc Technology Produced by the Sony Service Centre (Europe). 3rd edition, published 1995. Prepared by Sony’s technical staff, this is the best book on compact disc technology that we have ever come across. It covers digital audio in depth, including PCM adapters, the Video8 PCM format and R-DAT. If you want to understand digital audio, you need this reference book. 305 pages, in paperback at $55.95. Power Electronics Handbook Components, Circuits & Applica­ tions, by F. F. Mazda. Published 1990. Previously a neglected field, power electronics has come into its own, particularly in the areas of traction and electric vehicles. F. F. Mazda is an acknowledged authority on the subject and he writes mainly on the many uses of thyristors & Triacs in single and three phase circuits. 417 pages, in soft cover at $59.95. Surface Mount Technology By Rudolph Strauss. First pub­ lish-ed 1994. This book will provide informative reading for anyone considering the assembly of PC boards with surface mounted devices. Includes chapters on wave soldering, reflow­ soldering, component placement, cleaning & quality control. 361 pages, in hard cover at $99.00. Electronics Engineer’s Reference Book Edited by F. F. Mazda. First pub­ lished 1989. 6th edition 1994. This just has to be the best reference book available for electronics engineers. Provides expert coverage of all aspects of electronics in five parts: techniques, physical phenomena, material & components, electronic design, and applications. The sixth edition has been expanded to include chapters on surface mount technology, hardware & software design, Your Name__________________________________________________ PLEASE PRINT Address____________________________________________________ _____________________________________Postcode_____________ Daytime Phone No.______________________Total Price $A _________ ❏ Cheque/Money Order ❏ Bankcard ❏ Visa Card ❏ MasterCard Card No. Signature_________________________ Card expiry date_____/______ Return to: Silicon Chip Publications, PO Box 139, Collaroy NSW, Australia 2097. Or call (02) 9979 5644 & quote your credit card details; or fax to (02) 9979 6503. semicustom electronics & data communications. 63 chapters, in paperback at $140.00. Radio Frequency Transistors Principles & Practical Appli­ cations. By Norm Dye & Helge Granberg. Published 1993. This timely book strips away the mysteries of RF circuit design. Written by two Motorola engineers, it looks at RF transistor fundamentals before moving on to specific design examples; eg, amplifiers, oscillators and pulsed power systems. Also included are chapters on filtering techniques, impedance matching & CAD. 235 pages, in hard cover at $85.00. Newnes Guide to TV & Video Technology By Eugene Trundle. First pub­ lish-ed 1988, reprinted 1990, 1992. Eugene Trundle has written for many years in Television magazine and his latest book is right up date on TV and video technology. 432 pages, in paperback, at $39.95.  Title Price  Newnes Guide to Satellite TV  Servicing Personal Computers  The Art Of Linear Electronics  Optoelectronics: An Introduction  Digital Audio & Compact Disc Technology  Power Electronics Handbook  Surface Mount Technology  Electronic Engineer’s Reference Book  Radio Frequency Transistors  Newnes Guide to TV & Video Technology $55.95 $59.95 $49.95 $55.95 $55.95 $59.95 $99.00 $140.00 $85.00 $39.95 Postage: add $5.00 per book. Orders over $100 are post free within Australia. NZ & PNG add $10.00 per book, elsewhere add $15 per book. TOTAL $A February 1994  9 Manufacturer's Data On The ISD2590P Single-Chip Voice Recorder IC This second-generation series of solid-state audio ICs from Information Storage Devices features extended recording/playback times as well as a new pushbutton operation mode & lower distortion. By DARREN YATES Following close on the heels of the original ISD1000 series, Information Storage Devices has released the new second-generation of solid state audio devices - the ISD2500-series. The most notable feature of the new range is that the EPROM array has jumped in size from 128,000 to 480,000 bits, which has allowed the much greater recording times. The 2500-series comes in four versions, the ISD2545, ISD2560, ISD2575 and ISD2590 which have 45, 60, 75, and 90 seconds duration respectively. The frequency bandwidth for the devices range from 4.5kHz for the 2545 down to 2.3kHz for the 2590. Fig.1 shows the basic block diagram of the internals of the IC. As with the ISD1000-series, the new 2500-series uses a patented method of storing INTERNAL CLOCK ANA IN ANA OUT MIC MIC REF AGC AMP SAMPLING CLOCK TIMING ANALOG TRANSCEIVERS ANTIALIASING FILTER DECODERS XCLK PREAMP analog signals in EPROM cells. The technique is similar to programming an ordinary EPROM except that in this case, the cell isn't blasted with a high or low voltage level but in small increments. The output of the cell is compared with the input signal and while the cell output is below the sampled input, the device continues to incrementally charge up the cell. When the two are equal, programming of that cell ceases. The size of the incremental charges is such that there are 256 possible levels which is equivalent to a conventional 8-bit system, except that that this method requires only 1/8th the amount of storage elements for the same recording time. Looking at Fig.1, input signal is applied either to the MIC preamp or SMOOTHING FILTER SP+ MUX AMP 480K CELL NONVOLATILE ANALOG STORAGE ARRAY SP- AGC POWER CONDITIONING VCCA +5V VCCD +5V ADDRESS BUFFERS A0 A1 A2 A3 A4 A5 A6 A7 A8 A9 DEVICE CONTROL OVF PD P/R CE EOM AUX IN Fig.1: block diagram of the ISD2500 series analog voice recorder IC. The device stores the audio signal in an internal 480K EPROM that retains memory even when the power is switched off. 10  Silicon Chip directly to the main preamp via the ANA IN pin. From here, the signal undergoes automatic gain control (AGC) to produce the optimum recording level. After this, a 5-pole anti-aliasing filter removes the upper frequency signals. Transferring the signal directly to the cells is virtually the same as for the 1000-series with two rows of analog transceivers (or sample & hold circuits) which perform `parallel programming' of a given row of cells. At present, details are sketchy on the number of cells in each row but the system works with one row is receiving samples in real time while the other is programming multiple cells simaltaneously. When replaying, the stored signal passes through another 5-pole filter to remove components of the internal clock frequency, then fed through a multiplexer and out through the bridge amplifier. The internal clock does not require any external components, but it is also possible to use an external clock drive which is fed into the XCLK pin. Faster clocking The benefit of this is that by feeding the device with a higher than usual clock frequency, you can obtain some improvement in sound quality with corresponding sacrifice in recording time. However, you can't extend the recording time by decreasing the clock frequency. The reason is that if the clock frequency drops below the required level, the sampling is such that the filters can no longer remove the clock frequency component from the audio making it garbled and almost impossble to understand. Fig.2 shows a table of the various devices and their sampling rates, bandwidths and required external clock inputs. The EPROM array is divided up Table 1 Part No. Duration (secs) Input Rate Bandwidth Required XCLCK ISD2545 45 10.6kHz 4.5kHz 1365.3kHz ISD2560 60 8.0kHz 3.4kHz 1024kHz ISD2575 75 6.4kHz 2.7kHz 819.2kHz ISD2590 90 5.33kHz 2.3kHz 682.7kHz into 600 equal spaced sections, each of which can be accessed via the 10 address lines, A0 through to A9 (0 to 257 hex). For the ISD2560 60-second version, this gives a resolution of 0.1 seconds for each division, similar to the first series. The other addition is the new OVFbar (overflow) output. When in either record or playback mode, this line pulls low when the device has reach the end of its memory, or is as full as a boot. The benefit is that it is easy to cascade devices together and using this pin to control the next device in the chain. With regards to cascading devices, its possible to extend the recording time without limit. Using the EOM (end-of-message) and OVF lines, the first device is connected as the master and a number of other devices connected as `slaves' or memory modules. However, the cost of such a system is likely to be prohibitive. Pushbutton mode One of the more interesting features is the addition of a push-button mode. This allows the device to triggered by the rising or falling edge of signal rather than having to tie the corresponding input high or low. The makes design of peripheral circuitry much easier. The mode is entered into by pulling the two most significant address lines high as well as the M6 mode pin. The chip enable (CE) pin now becomes a toggle START/PAUSE control while the power down (PD) line is now a STOP/RESET control. The pause feature is a very useful one as it allows you to stop recording or playback of a message, and then to continue on from that spot, just as you would with a normal tape deck. When recording, pressing the PAUSE key inserts an EOM (endof-message) marker at the present memory location. When replaying, each time, the IC comes across the EOM marker, it pauses at that memory location. Pressing the START/PAUSE key will cause the IC to begin playing the next message starting at the next memory location. This is ideal for example if you have five commands which explain how a piece of machinery should be used. At the end of each command, the user has to press the START/PAUSE key to hear the next command. The pause prevents the user from hearing all five commands at once and possibly making errors. Message looping There are many applications where you would record a message into the device and then have it continuously loop, playing the message continuously. Examples of this would be answering machines, in-store advertising, etc. By pulling the M3/A3 address line high, the device enters the Message Looping mode. It is activated when the Chip Enable (CE) line is pulled low. This continous looping continues until the CE line is pulled low again at which time the current mode and address lines are looked at and the corresponding mode executed. The ISD2500-series are still fabricated in the same 28-pin DIL package but are also available in SOIC, TSOP and bare die formats. In addition, theses devices are also avilable in a low-voltage range (3.6-4.0V). The total harmonic distortion for all devices is quoted as 1% <at> 1kHz and the output power amplifier can supply 50mW into 16W. If using an 8W speaker, a 10W 0.25W resistor should be placed in series. The output stage is a bridge amplifier with both anti-phase signals appearing at pins 14 and 15. To connect the device to an external amplifier, a series capacitor and 10kW resistor or pot should be connected to one of the outputs while the other is left floating. Connecting either output to ground will more than likely destroy the output stage. See the project based on the ISD2590P on page 16 for an SC example of this. VCR ALIGNMENT TOOL KIT • 7 Assorted head & guide aligners • Hex key set • Retaining ring remover • 3 Reversible screwdrivers – SML – Flat – Philips • Spring hook • Fitted vinyl • Micro screwdriver • Zippered • VCR head puller Our Low Price $99.95 WOMBAT COMPONENTS WOMBAT COMMUNICATIONS 83 - 85 Railway Ave Werribee, Vic 3030 Phone: (03) 742 7330 Fax: (03) 741 6834 AUDIOPHILES! Now high audiophile quality components & kits are available in Australia. Buy direct & save. *Kimber, Wonder, Solen & MIT Capacitors *Alps Pots *Holco resistors *High Volt. Cap *Gold Terminals & RCA *WBT Connectors *Kimber Cables *Interconnect Cables *Output Transformers (standard or customised) *Power Transformers *Semiconductors *Audio Valves & Sockets *Wonder Solder *Wetborne Labs Accessories Valve & Solid State Pre-Power Amplifier Kits *Contain Stereo 80 Valve Power Amp (As per Elect. Aust. Sept. & Oct. ‘92) *Welborne Labs Hybrid Preamp. & Solid State Power Amplifier Send $1.00 for Product Catalog. PHONE & FAX: (03) 807 1263 CONTAN AUDIO 37 WADHAM PARADE MT. WAVERLEY, VICTORIA 3149. February 1994  11 SILICON CHIP If you are seeing a blank page here, it is more than likely that it contained advertising which is now out of date and the advertiser has requested that the page be removed to prevent misunderstandings. Please feel free to visit the advertiser’s website: dicksmith.com.au SILICON CHIP If you are seeing a blank page here, it is more than likely that it contained advertising which is now out of date and the advertiser has requested that the page be removed to prevent misunderstandings. Please feel free to visit the advertiser’s website: dicksmith.com.au SILICON CHIP If you are seeing a blank page here, it is more than likely that it contained advertising which is now out of date and the advertiser has requested that the page be removed to prevent misunderstandings. Please feel free to visit the advertiser’s website: dicksmith.com.au SILICON CHIP If you are seeing a blank page here, it is more than likely that it contained advertising which is now out of date and the advertiser has requested that the page be removed to prevent misunderstandings. Please feel free to visit the advertiser’s website: dicksmith.com.au Build this 90-second message recorder If the 16-Second Message Recorder published in July 1993 wasn’t long enough for you, then try this 90-second model. It runs from a 6V battery & features more power output, a pause but­ton, 90 seconds of continuous recording time & zeropower memory storage. By DARREN YATES There’s no doubt about it – solid state audio recorders are the big noise in electronics at the moment. This was shown by the popularity of our 16-Second Message Recorder project published in the July 1993 issue of SILICON CHIP. So popular was this project that it spawned a couple of pre-built imported surface-mount modules and at least one retail­er is now stocking the device 16  Silicon Chip as a regular catalog item. But as the 286 PC was to the XT, so is this new 90-second sound recorder to that original project. It’s based on the sec­ ond-generation of sound recorder ICs just released by Information Storage Devices. Called the ISD2500-series, there are four mem­ bers each containing 480,000 EPROM cells as opposed to the 128,000 in the ISD1000-series. Despite the popularity of the original design (or perhaps because of it), there were quite a few calls asking “how can you make it longer?” It seems as though people these days leave lots of messages on the fridge! This 90-Second Message Recorder uses the new ISD2590P voice storage IC. It operates from a 6V battery and includes a PAUSE/START key and a separate power amplifier IC. Looking at the IC briefly, instead of using standard digi­tal technology, the ISD2590P uses a patented analog method which allows analog voltages to be stored directly into the EPROM cells. It contains everything to make a complete audio record\playback system from microphone preamplifier to AGC, 480K EPROM storage cells as well as anti-aliasing filters and output amplifier. During recording, this device samples the incoming audio signal and D1 1N4004 0.1 2.2k 0.22 ELECTRET MIC 0.22 S4 +6V 10 10k PARTS LIST 220 16VW 28 7 9 10 16 17 VCCD A6 A8 A9 VCCA 14 10k MIC SP+ 18 MIC REF A OUT 21 6V 10 1k VOLUME 10k LOG 1 3 6 2 4 1k +6V RESET REWIND S2 23 24 100k RECORD R/P AGC EOM 4.7 A0 26 A2 XCLK A3 A7 A5 A4 12 13 8 B A +6V 22k C B E 22k A1 E C VIEWED FROM BELOW 27 6 Q1 BC548 B 25 1 2 8W PLAY S3 PD 0.1 10 20 CE START PAUSE S1 19 470k A IN IC1 ISD2590P 100k 470 5 IC2 LM386 E C LED1 PLAY  LED2 RECORD Q2 BC558  3 4 680  680  5 4.7k K B C E Q3 BC548 90-SECOND MESSAGE RECORDER Fig.1: the circuit is based on IC1, an ISD2590P 90-second voice storage IC. Its output appears at pin 14 & is fed to an LM386 audio amplifier (IC2) which in turn drives a small loudspeaker. Transistors Q1-Q3 drive the PLAY & RECORD indicator LEDs (LED 1 & LED 2). stores these samples as analog voltages in the EEPROM. This technique is eight times more efficient than current digital technology and has the added bonus of zero power for memory retention. In fact, ISD guarantee that it will hold a message for 100 years. And since the writing cycle is much more gentle than the usual digital EPROM programming methods, you can achieve up to 100,000 record cycles with the device. For more details on this device, take a look at the data article published elsewhere in this issue. Operation OK, let’s now go through the operation of the Message Recorder. Initially, when power is applied, nothing will appear to happen. If you now set the PLAY/RECORD switch S3 to RECORD, the unit is ready to record. Pressing the START/PAUSE button S1 once will start the device recording and LED 2 will light up. Recording will continue until either the device runs out of memory or you press either the START/PAUSE button or the REWIND/ RESET button. Pressing the START/PAUSE button will stop recording but will keep the address counter at its present position – it works just like the PAUSE button on your tape deck. Pressing the RESET/ REWIND button will also stop recording but will reset the address counter back to zero. Pressing the START/ PAUSE button again will commence recording from the beginning, erasing any previous recording. To play back what you have just recorded, flick switch S3 into PLAY mode and press the START/PAUSE button. You will now hear the first recording which will continue until either 90 seconds has passed or until the device comes up against an endof-message indicator. At this point, the device goes into an automatic ‘pause’ mode, and by pressing the START/PAUSE button again, you will hear the next recording. At any time, you can PAUSE the 1 PC board, code 01202941, 97 x 85mm 1 battery clip 1 6V battery holder 4 AA size cells 1 red snap-action pushbutton switch 1 green snap-action pushbutton switch 2 SPDT toggle switch 4 10mm tapped 3mm spacers 1 electret microphone insert 1 28-pin machined IC socket 4 PC stakes 1 8Ω 250mW loudspeaker 1 knob Semiconductors 1 ISD2590P 90-second audio recorder (IC1) 1 LM386 low-power audio amplifier (IC2) 2 BC548 NPN transistors (Q1,Q3) 1 BC558 PNP transistor (Q2) 1 5mm green LED (LED1) 1 5mm red LED (LED2) 1 1N4004 rectifier diode (D1) Capacitors 1 470µF 16VW electrolytic 1 220µF 16VW electrolytic 2 10µF 16VW electrolytic 1 4.7µF 25VW electrolytic 1 1µF 50VW electrolytic 2 0.22µF 63VW MKT polyester 2 0.1µF 63VW MKT polyester Resistors (0.25W, 1%) 1 470kΩ 1 2.2kΩ 2 100kΩ 2 1kΩ 2 22kΩ 2 680Ω 2 10kΩ 1 10Ω 1 4.7kΩ Miscellaneous Screws, washers, solder, tinned copper wire. playback by pressing the START/ PAUSE button or reset the device to the begin­ning by pressing the RESET/ REWIND button. Circuit details Let’s take a look then at the circuit diagram in Fig.1. As you can see, there are just two ICs, the ISD2590P and an LM386 audio amplifier IC. The latter February 1994  17 10uF 10k 0.22 22k 22k 1k 0.1 1uF Q1 Q2 680  0.22 470k 220uF MIC IC1 ISD2590P 680  4.7uF 10k 10uF A LED1 A LED2 IC2 386 1 1 4.7k 2.2k Q3 1k 0.1 D1 10  S1 470uF S2 VR1 100k BATT SPKR S3 100k Fig.2 (above): install the parts on the PC board as shown here. Use a socket for IC1 & note that Q2 is a PNP transistor while Q1 & Q3 are both NPN types. Note also that switches S1 & S2 are oriented with their flat edges towards IC1. Fig.3 at right shows the full-size etching pattern for the PC board. IC is used to boost the 2590P’s output signal. Looking at the circuit, the input signal is obtained from an on-board electret microphone insert, which is biased via the 2.2kΩ and 10kΩ resistors. The 10µF capacitor at the junction of these two resistors provides supply decoupling and prevents clock hash from IC1 entering the audio stage. As soon as power is applied, the circuit is switched to a special ‘push-button’ mode by virtue of the fact that address lines A6, A8 and A9 are tied high. The CHIP ENABLE (CE) pin becomes the START/PAUSE control line (pin 23) and the POWER DOWN (PD) pin becomes the RESET/REWIND control. Because these controls are now edge-triggered, only pushbutton switches are required. The AGC (automatic gain control) filter components are the 470kΩ resistor and the 4.7µF capacitor on pin 19. Replay and record selection is made via switch S3. By pulling the R/P input at pin 27 low, the device is placed in record mode and when it is high, it’s in play mode. Switch S3 also controls the two LEDs which display the operating mode. With switch S3 low, transistor Q1 is biased off but Q2 is turned on. With switch S3 high, Q2 is biased off but Q1 is turned on. However, both LEDs are also controlled by transistor Q3, which is driven by the EOM output at pin 25 via a 4.7kΩ resistor. When the START/ PAUSE button is pressed, the EOM line is pulled high for the duration of the first message. This is always the message that begins at address location 0 hex. While the EOM line is high, either LED 1 or LED 2 will light up depend- ing upon the operating mode – LED 1 for PLAY and LED 2 for RECORD. If the START/PAUSE or RESET/REWIND buttons are pressed while the device is either currently recording or playing back, the current operation ceases and the corresponding LED goes out. Since we are using the internal microphone preamplifier, the output which appears at pin 21 must be recoupled back into the main preamplifier stage whose input is at pin 20. This is done via a 1µF capacitor and 1kΩ resistor. Output signal The output at pin 14 is coupled via a 10kΩ resistor and 10µF capacitor to a 10kΩ volume control pot. The 10kΩ series resistor is included to improve RESISTOR COLOUR CODES ❏ ❏ ❏ ❏ ❏ ❏ ❏ ❏ ❏ ❏ No. 1 2 2 2 1 1 2 2 1 18  Silicon Chip Value 470kΩ 100kΩ 22kΩ 10kΩ 4.7kΩ 2.2kΩ 1kΩ 680Ω 10Ω 4-Band Code (1%) yellow violet yellow brown brown black yellow brown red red orange brown brown black orange brown yellow violet red brown red red red brown brown black red brown blue grey brown brown brown black black brown 5-Band Code (1%) yellow violet black orange brown brown black black orange brown red red black red brown brown black black red brown yellow violet black brown brown red red black brown brown brown black black brown brown blue grey black black brown brown black black gold brown the loading of IC1’s output stage. The volume control feeds IC2, an LM386 audio amplifier IC. This is connected in its minimum-component mode and has a gain of 20. The output from the LM386 is approximately 300mW into an 8Ω loudspeaker with the 6V supply. Power is provided from a 6V battery, with four AA cells being the most appropriate. Diode D1 provides reverse-polarity protection and the 220µF capacitor provides supply decoupling. Typically, current consumption should be about 6-8mA quiescent and about 30-35mA when recording or replaying. Construction All of the components for the 90-Second Message Recorder, except for the battery and power switch, are installed on a PC board measuring 90 x 97mm and coded 01202941. Before you begin any soldering, check the board thoroughly for any shorts or breaks in the copper tracks. These should be repaired with a small artwork knife or a touch of the soldering iron where appropriate. Next, you should make sure that the components will fit into the holes drilled. You will probably have to do a little work for the mounting of the volume control and the PLAY/RECORD switch. You can use a 3mm drill for the volume control hole and then enlarge it with a tapered reamer or round file to suit. Once you’re happy that everything is correct, start off by installing the wire links. Use the overlay wiring diagram (Fig.2) to make sure that they go into the correct locations, then install the resistors, capacitors, diode and transistors. Note that most of these components are polarised and need to be installed the correct way around for the circuit to work. Again, use the overlay wiring diagram to make sure that everything is correct. Because the ISD2590P is an expensive device to replace, we suggest that you use a 28-pin machined IC socket – not one of the cheaper variety. The cheap ones have a habit of becoming unreli­able after a very short time. Solder the IC socket in the same way you would the IC. You’ll find that the socket has a notch in one end, just as the IC does. This makes it easy to remember which way around the IC must be plugged in if it ever needs to be removed. Next up, solder in the LM386 amplifier IC. Once that has been done, you can install the switches. All of these except for the power switch S4 are installed on the board. The two snap-action switches should fit snugly into position on the board. Make sure that the flat section on these switches if facing towards the top of the board (ie, towards IC1). Testing Finally, insert the ISD2590P into the 28-pin socket. Make sure that it goes in the right way around. This done, connect a 6V battery in series with an external power switch and your multimeter. When the power is switched on, you should find that the current consumption is about 8-10mA. If it’s any more than 15mA, switch off immediately and check the board for possible solder shorts or component positioning errors. If everything appears to be in order, follow the operating routine outlined earlier to record and play back to your SC heart’s content. Subscribe now to the largest faults & remedies library in Australia ✱ ✱ 1994 manuals are now available. Our database is regularly updated with information supplied by technicians such as yourself. ✱ Exclusive backup service by qualified technicians. ✱ ✱ Over 10,000 faults and remedies on file with flow charts and diagrams. Covers Colour TVs and VCRs of all brands sold in Australia EFIL Phone or fax now for your FREE information package ELECTRONIC FAULT INFORMATION Reply Paid 4 P.O. Box 969 AIRLIE BEACH 4802 Ph 079 465690 Fax 079 467038 February 1994  19 CIRCUIT NOTEBOOK Interesting circuit ideas which we have checked but not built and tested. Contributions from readers are welcome and will be paid for at standard rates. Using two train controllers to operate one section This circuit allows you to control one track section (block) of a model railway with two IR controllers (as described in the April, May & June 1992 issues of SILICON CHIP). Relay contacts direct the output of each controller to the track so that only one controller is connected to the track at one time. Selection of each controller is made using a momentary auxiliary output from each. Throttle 1’s auxiliary output con­nects to pin 1 of IC1a and Throttle 2’s auxiliary output connects to pin 6 of IC1b. Nor gates IC1a and IC1b are connected as an RS flipflop. When pin 1 goes momentarily high, pin 4 goes high and pin 3 low. Conversely, when pin 6 goes high, pin 3 is high and pin 4 low. These outputs remain in this state until triggered again by a high pulse on pins 1 or 6. The output at pin 3 drives transistor Replacing selenium cells with solar cells Photographers have used Weston Master lightmeters for many years. However, the selenium cell degrades with time and is no longer obtainable. I have found the following arrangement quite satisfactory. I used a silicon solar cell obtainable Digital tachometer & dwell angle meter For years to come there will be a large number of cars that rely on conventional ignition. When the engine is being tuned, one of the most important steps is to ensure that the points are set correctly. This circuit has been designed to meet that re­quirement. The signal from the distributor points passes via diode D2 and an RC filter to the base of Q1. The resulting 20  Silicon Chip +12V 10k FROM AUX OUTPUT THROTTLE 1 +12V 10k FROM AUX OUTPUT THROTTLE 2 4001 14 11 IC1a 2 5 6 IC1b 3 10k Q1 BC327 TO THROTTLE 1 TRACK VOLTAGE 1k 4 D1 1N4004 RLY1 LED1 RED  7 RLY1a RLY1b +12V 10k Q2 BC327 1k RLY2b D2 1N4004 RLY2 LED2 GREEN  TO TRACK RLY2a TO THROTTLE 2 TRACK VOLTAGE Q1 and relay RLY1 which connects Throttle 1 to the track. Similarly, the output at pin 4 drives transistor Q2 and relay RLY2 to connect Throttle 2 to the track. The controlling throttle is indicated by a LED which is activated whenever its relay is powered. A red LED is used for Throttle 1 (LED 1) and a green LED for Throttle 2 (LED 2). Alf McKeon, Browns Plains, Qld. ($25) from Tandy (Cat No 276-124) which measures 20 x 40mm and fits in the recess provided in the meter for the selenium cell. Because silicon has a very high response in the near infrared compared to selenium, I used two layers of a cyan 50 printing colour filter cut to fit the cell recess (cyan filters come in colour printing filter sets but are rarely used). Then the new cell is fitted and a foam plastic used to press it and hold it in place. Although smaller in area than the selenium cell it replac­es, its sensitivity is much higher and with the two layers of filter, it equals the original cell for normal light condi­tions. V. Erdstein, Highett, Vic. ($15) square wave at Q1’s collector is inverted by IC1a and fed to the clock input of IC4, a 4017 divide-by-10 counter. The negative edge of the output pulse at pin 11 triggers monostable IC3 which goes high for 0.5 seconds, on every 10th pulse. In tacho mode, this becomes a time reference for IC8, a 74C925 4-digit counter and display driver. IC3’s output at pin 3 is coupled via 470pF capacitors to pins 5 and 12 of IC8 and thus provides the latch enable and reset signals for IC8. For dwell angle measured in degrees of rotation, the 0.5-second time slot is extended and serves as an updating time delay. From IC1a, the negative going ignition pulse is inverted by IC1b before being fed to pin 14 of IC5, a phase lock loop IC which is controlled by IC6, the latter connected to divide by 6, 4 or 3 for 4, 6 and 8-cylinder engines. To measure the dwell angle, IC6 and IC7, a binary divide-by-15 D1 1N4004 +12V FROM BATTERY IN 470pF OUT 7805 +5V GND 470 0.1 0.1 100uF 50k 16 1k Q1 BC337 D2 1N4007 100k FROM POINTS 50k .01 14 14 IC1a 4009 15 16 6 IC1b 16 11 12 11 1k 470pF 7 14 13 8 1k 15 IC5 4046 4 14 300k 10k 0 IC6 4017 CLK 6 4 9 3 50k 11 5 R 8 8 3 15 10k 1.5 TANT 1k VR1 500k +5V 14 4 16 11 2 S1c S1b 16 1 2 11 11 6 7 IC1d IC1c 2 1 10k 6 IC2c 4 11 11 LE IC8 74C925 CLK 5 470pF 12 12 10k 7 CLK QD 10 IC7 40161 RC LO IB IC 4 ID 5 8 6 R A 8 B 6 C 7 S1d 2 16 470pF 55 9 3 CLR IA EN 120  +5V S1a 2 100 1 10 1 3 S2 2 1 15 9 7 +5V 1.5 TANT 15 13 8 11 1 3k 3 IC3 555 6 IC4 4017 8 7 5 1 2 +5V 2 13 7x 82  11,16 11,16 14 15,10 15,10 15 3,8 1 2,6 2 1,5 3 18,12 18,12 4 17,7 D 9 10 9 DP a b a c d e f e g b d c f DISP1 HDSP5523 g Q5 BC337 14 DISP2 HDSP5523 13 14 13 Q4 BC337 Q3 BC337 S1 : MODE 1 : TACHOMETER 2 : DWELL ANGLE S2 : ENGINE 1 : FOUR CYLINDER 2 : SIX CYLINDER 3 : EIGHT CYLINDER 5 IC1f 4 4011 8 IC2b 9 counter, are connected so that IC5 multiplies each full igni­tion cycle of a 4- cylinder engine by 90, a 6-cylinder engine by 60 and an 8-cylinder engine by 45. These pulses are then gated with the ignition pulses from +5V 14 7 10 12 IC2a 11 3 IC1e Q2 BC337 2 13 IC1b by IC2c and fed to pin 8 of IC8, the 4-digit counter. The number displayed corresponds to dwell angle in degrees. Inverters IC1e & IC1f plus gates IC2a & IC2b suppress the leading zero for readings below 1000 RPM and when in dwell angle mode. Dual 2-digit HDSP 5523 displays were used, as these can be read in full sunlight. K. Benic, Forestville, NSW. ($40) February 1994  21 Instrumentation programming – doing it the graphical way In the past, PCs have been used to control data acquisi­tion and test equipment via programs written in the conventional way, with hundreds or thousands of lines of text based code. This article discusses LabVIEW – software based on graphical program­ming. By JACK BARBER The introduction in 1986 of Lab­ VIEW (Laboratory Virtual Instrument Engineering Workbench) for the Macintosh revolution­ ised PC-based instrumentation with the concept of graphical programming – developing block diagrams rather than writing conventional, text-based code. Lab­ VIEW was the first graphical program- ming language used to integrate several popular classes of instrumentation hardware for test and measurement applications. In 1992, National Instruments announced LabVIEW version 2.5 for Sun SPARCstations while version 3.0, introduced in 1993, made graphical instrumentation applications complete­ ly portable between Macintoshes, Win­dows PCs or Sun SPARCstations. This article explains the benefits of graphical program­ming with LabVIEW and the characteristics and features of LabVIEW that differentiate it from other products that appear to have a similar look and feel. Graphical programming Graphical programming offers the ability to create software applications to those who otherwise do not have the time or skills to program using conventional text languages. Graphical programming lets the user draw a diagram or a picture to explain a process or algorithm. A user can easily scan a picture of a graphical program for relevant features, data flow structure and complex relationships that would otherwise be hidden in the code of a text-based program. Graphical programming can be tailored for a particular application area. By supplying the user with familiar tools and terminology, the software package serves as an enhancement rather than a hindrance to the application. A LabVIEW program or subpro­gram is like an instrument with front-panel controls. The “in­strument” measures inputs and displays outputs. This instrument also has internal circuits. G, the “language” in LabVIEW, gives users the ability to draw the schematic, so to speak, for these circuits. These software emulations of hardware instruments are therefore called “virtual instruments,” or VIs. Graphics vs icons Fig.1: system developers use pull-down and pop-up menus to equip the front panel with indicators and controls. The front panel serves as the graphical user interface during program execution. 22  Silicon Chip Today, several software products use icons for visual representations however few of these are true graph- ical program­ming systems. Most are menu-driven systems where each icon repre­sents a function and contains a list of options. Users connect these icons to specify an action. Icon-based systems are typical­ ly limited by a small set of functions, options and ways in which users can connect the icons. However, such programs may well satisfy users who have simple application requirements that will not become more demanding in the future. With Lab VIEW ’s graphical programming approach, the novice can quickly assemble simple programs such as those typically created with menu-driven packages. However, more experienced users will find LabVIEW also offers a good alternative to con­ven­tional text-based languages such as BASIC or C. Like convention­ al programming systems, LabVIEW incorporates features such as hierarchy, execution control, programming structures and also a compiler. Virtual instrument Before selecting software, it is important to consider how you want your system to present data. Due to limited space on the screen, combining the GUI elements with the functional elements in a diagram (as some software products do) is impractical for complex applications. A LabVIEW VI has separate panels and a diagram optimised for operating and programming, respectively. On the front panel, users arrange the controls and indicators in a logical order, add background pictures and create custom controls to add context to the GUI (Graphical User Interface). In the diagram, small graphi­cal equivalents of the GUI elements save space and make it easier to construct a block diagram. Fig.2: Temperature System VI has a While Loop that contains a For Loop (which acquires a group of temperature readings) and a Case Structure (which determines if the data is to be analysed.) You can click a switch, move a slider, tweak a knob, or type a value on the front panel to interactively control the system during execution. Meanwhile, the indicators provide feed­back and results. LabVIEW can store the data by printing the front panel or by saving it as a picture file. LabVIEW’s block diagram defines what the virtual instrument (VI) does. The block diagram contains terminals (smaller repre­sentations of the front panel controls and indicators) that pass data to and from the front panel. You connect these terminals using the wiring tool to pass data from one block to the next. The diagram may have multiple data paths and thereby sim­ultaneous operations. The LabVIEW system also has functional blocks to perform simple arithmetic functions, advanced Creating the front panel In LabVIEW, you first create the front panel to define the input and output parameters of the program. LabVIEW has controls and indicators (knobs, sliders, switches, LEDs, text boxes, charts and graphs) in hierarchical menus. Once an indicator or control is selected and placed on the panel, it can be moved, sized, labelled and configured in terms of data type, dimension, range, default and mechanical action. The user can import pic­ tures and controls to tailor a panel to a specific application. Fig.3: LabVIEW is a graphical programming system for developing data acquisition & instrument control applications on Macintosh computers, Windows PCs & Sun SPARCstations. February 1994  23 distribute VIs to users who can load and run VIs but cannot edit them or display their diagrams. This protects the propriety rights and integrity of VIs. The LabVIEW Run-Time System can serve as a low-cost test station or as an efficient way to package and resell VIs. Input/output (I/O) Most applications require the use of hardware for data collection, so users should also consider what types of I/O their application will require –plug-in data acquisition and/or in­ strument control. The software should work with a variety of hardware – it is then easy to integrate different types of hard­ware into one system. The data acquisition (DAQ) hardware should have ready-to-use instrument drivers available and it should be easy to add new drivers. LabVIEW has drivers for more than 300 GPIB, VXI and RS-232 instruments. The drivers consist of high-level functions with a front panel to operate each instrument. More importantly, each icon can be incorporated into a block diagram with other driver icons to build a complete test system. Data analysis Fig.4: Lubrizol Corporation in Wickliffe, Ohio (USA) uses a Macintosh Computer running LabVIEW in its high temperature fluid durability cycling tests. Lubrizol uses LabVIEW to create unique screens to easily acquire, analyse and save raw data from the tests. This process has eliminated many of the variables involved in the analysis of the data & streamlined the report generation process. acquisition and analysis routines and file I/O and network operations that store or re­trieve data in ASCII, binary or spread­sheet formats. LabVIEW also contains a formula node for typing in simple arithmetic equa­ t ions. For more complicated routines, the Code Interface Node (CIN) links external code to the diagram. This feature is import­ant for users that have already developed routines, like analysis algorithms, in a conventional language. LabVIEW includes extensive tools to develop, test and debug a VI system. The Help window describes each VI and its connec­ tions. The program immediately indicates incorrect wire connec­ tions with a dashed line. In addition, the Error window lists syntax errors. Execution highlighting traces the data paths during VI execution. The single-step mode and breakpoints 24  Silicon Chip also aid in VI debugging. LabVIEW has programming structures such as for loops, while loops and case statements for sequential, repetitive and branch­ing operations that determine if or how many times a set of functions will be executed. Graphical compiler LabVIEW is the only software of its type that features a graphical compiler – a system that compiles its block diagrams into machine code. This produces programs that execute at speeds comparable to compiled C programs. Consequently, LabVIEW programs execute 10 to 1000 times faster than those of any other graphi­cal instrumentation programming system. The graphical compiler also creates VIs for the LabVIEW Run-Time System. With this compact, low-cost version of LabVIEW, system developers can Users need to convert acquired data into meaningful re­sults. The Analysis VI libraries offer digital signal processing (DSP), digital filtering, statistics and numerical analysis functions. Also included are functions for array manipulation, complex arithmetic and statistical functions, Fast Fourier Trans­­ form (FFT) and Fast Hartley Transform (FHT) integration, differ­ entiation, convolution and correlation, power spectrum and pulse parameters; finite impulse response and infinite impulse response digital filters; win­ dowing functions; signal generation; linear, exponential and polynomial curve fitting; advanced statistics; and complex and matrix operations. As you can see, LabVIEW is a comprehensive solution to virtual instrument programming. It is intuitive and the resultant compiled programs run very fast. For further information on instrumentation programming and other data acquisition products, contact Tony O’Donnell, National Instruments Australia Corporation, PO Box 466, Ringwood, Vic 3134. Phone (03) 879 SC 9422 or fax (03) 879 9179. Light, compact & efficient 12-240VAC 200W inverter This light & compact 200W 12V-240VAC inverter can drive mains appliances, including power tools, fluorescent & incandescent lights, TVs, etc from a 12V battery. It is ideal when camping, for use at building sites or as part of a solar power installation. By JOHN CLARKE This 200W inverter covers the medium power range and is suitable for powering household appliances such as power tools, hifi and video equipment and personal computers. It is unsuitable for driving microwave ovens, washing machines and other higher power appliances. While inverters described in electronics magazines in the past have usually employed heavy mains trans26  Silicon Chip formers (apart from our 2kW sinewave inverter), this new design uses a high frequency transformer which is small, light and efficient. To give a comparison, the 40W 50Hz square wave inverter published in the February 1992 issue of SILICON CHIP weighed about 1.25kg. This new design, which puts out five times as much power, weighs 1kg. Because it doesn’t use a mains trans- former, the new design also draws a much smaller current when in the standby condition; ie, when powered up but with no load connected. Its standby current is 55 milliamps which compares very favourably with the 1 amp standby current of the 40W inverter referred to above. Square wave The output waveform of the new inverter is a “modified square wave” with a duty cycle of 35%, the best compromise waveform for a low cost inverter. This is explained by the diagram of Fig.2 which shows the three Top of page: the 200W inverter is fitted with a low-profile 240VAC power point & is suitable for powering many power tools & other domestic applianc­es. ISOLATED VOLTAGE FEEDBACK +340V R1 +12V Q3 ISOLATED GATE DRIVER +12V Q1 T1 Q5 AC X Q2 Q6 ISOLATED GATE DRIVER 100 385VW AC Y Q4 R2 25kHz SWITCHMODE DRIVER ISOLATED GATE DRIVER 240VAC OUTPUT ISOLATED GATE DRIVER 0V OVERCURRENT AMPLIFIER Ri MODIFIED SQUARE WAVE GENERATOR DC-DC CONVERTER SQUARE WAVE 'H' PACK Fig.1: block diagram of the 200W inverter showing the high fre­quency DC-DC step-up stage & H-pack output stage. Fig.2: various 50Hz inverter output waveforms. (a) is the ideal; (b) has low amplitude; and (c) is the modified square wave output used in the 200W inverter. common inverter waveforms. Note that they all have the same RMS value of 240V. The sinewave is the ideal waveform since it has no harmonics and it swings over a range of ±340V peak. Sinewave output is usually reserved for high power inverters because of the extra complexity. The second common inverter waveform is the square wave which, despite having the required 240V RMS value, has a peak swing of only ±240V. This is • • • • • • • • Features often insufficient for correctly power­ ing appliances which rely on the peak voltage of the 50Hz mains waveform. This includes any appliance with a rectifier and filter capacitor power supply such as computers, VCRs, TV sets, hifi systems and so on. Then there is the “modified square wave”. There are many types of modified square wave inverters. Some start off with a low duty cycle and a high peak voltage (as in Fig.2c) on light loads and increase the duty cycle to a full square wave (Fig.2b) when driving a full load. This duty cycle variation is used as a means of Small size (1kg mass) Low standby current Modified square wave output Peak-peak voltage equal to mains sine wave Under voltage shutdown 30A over-current limiting Fuse protection Fully isolated output for safety Specifications Input voltage .......................................11-14.8VDC (12V lead acid battery) Output voltage ............................................ 240VAC modified square wave Power rating ....................................... 200W short term, 150W continuous Surge power .......................................................................................350W Standby current ..................................................................................55mA Full load current .........................................................25A DC (200W load) Output regulation .................................................................................< 8% Efficiency ................................................................ > 70% for loads > 60W 50Hz accuracy .....................................................................................±5% Fig.3: this diagram shows how the gate signals to the H-pack Mosfets are arranged to give the modified square wave output. February 1994  27 output volt­age regulation. However, it also means that the peak voltage will depend upon the load which is less than ideal. 28  Silicon Chip Our new 200W Inverter provides a fixed 35% duty cycle re­gardless of load current so that the peak voltage is maintained. Output regulation is achieved by keeping the peak voltage con­stant. Fig.1 shows the block diagram of the 200W Inverter. It incorporates a high frequency DC-DC converter and an H-pack output stage. The DC-DC converter has a switch­ mode driver to control Mosfets Q1 and Q2. These devices drive transformer T1 in push-pull mode. The step-up ratio is 38:1 and the resulting AC voltage is rectified by a full wave bridge Fig.4: the circuit of the 200W inverter. At left is the 25kHz DC-DC step-up section involving transformer T1. At top right is the H-pack output stage, while at bottom right is the 1MHz burst circuitry. February 1994  29 The pencil in this shot is pointing to Mosfet Q1. Q1 & Q2 are BUK436-100A Mosfets rated at 33 amps, 100 volts & 125 watts. They are mounted on the heatsink as shown in Fig.8. and filtered with a 100µF 385VW capacitor. The isolated feedback circuit adjusts the Mosfet switching so that the DC voltage from the inverter is maintained at +340V regardless of the load current. The Mosfets are protected against overcurrent if, say, an excessive load is connected to the inverter. Over­current protec­ tion is achieved by detecting the voltage drop across resistor Ri. If the voltage exceeds a preset level, the switchmode driver reduces the duty cycle applied to the Mosfets and thus reduces the overall current. You might think that the transformer step-up ratio of 38:1 is far greater than necessary to give the 340V required. This ratio has been made larger to offset inevitable losses in the inverter and to provide good output voltage regulation. The 340V supply rail is fully floating with respect to the 12V battery terminals by virtue of the step-up transformer and the isolated voltage feedback. This will prevent the battery terminals from being at a high potential above ground should a fault occur in any equipment powered by the inverter. Across the 340VDC supply are connected four high voltage Mosfets in an H-pack arrangement. Q3 is in series with Q4 while Q5 is in series with Q6. The junction between Q3 and Q4 is point X, while the junction between Q5 and Q6 is point Y. If Q3 is turned on and Q4 off, point X is pulled up to +340V. Conversely, if Q4 is on and Q3 off, then point X is pulled down to 0V. Similarly, point Y can be pulled down to 0V when Q6 is turned on and pulled up to +340V when Q5 is on. The square wave generator circuitry has four outputs which drive Q3, Q4, Q5 and Q6. This allows the circuitry to pull point X to +340V and point Y to 0V for one half of the 50Hz waveform, then pull X to 0V and Y to +340V for the other half of the 50Hz waveform. Note that the Mosfets are switched on for only 70% of the time so the overall duty cycle of the waveform is 35%. Fig.3 shows the switching process in the H-pack output stage. Each of the output Mosfets is a FRED FET (Fast Recovery Epitaxial Diode Field Effect Transistor), made by Philips. The term “FRED” means that they incorporate a fast recovery reverse diode which protects the device from peak reverse voltages which can be generated when driving inductive loads. Apart from incandescent lamps and heaters, virtually all mains appliances can be regarded as inductive. Circuit description The full circuit for the 200W Inverter is shown in Fig.4 While there is a fair amount of componentry involved, the basic circuit operation is the same as detailed in the block diagram. At the heart of the DC-DC converter is IC1, a TL494 pulse width modulation 30  Silicon Chip ▲ This photo highlights the 1MHz gate drive circuitry for the H-pack Mosfets. Note the tiny toroids which are wound as transform­ers. Fig.5 (facing page): the full wiring diagram of the inverter. Note the differ­ent diameters of enamelled copper wire specified for the links. Be sure to use heavy-duty cables where specified (see text) & take care with the orientation of transformer T1. REAR PANEL CORD-GRIP GROMMETS EARTH BLACK RED 1.25mm ENCU D2 10  10  ZD1 2.2uF 100V 1 2.2uF 100V Q5 ZD2 ZD3 TP2 ZD4 T1 Q7 D10 1 100pF D12 220k TP1 IC2 4050 D9 ZD5 Q8 100pF D14 220k 0.1 T4 1 150k D8 T5 T6 3.3k Q13 Q14 0.1 56k 820  IC4 IL300 220k 560pF 560pF 100uF 385VW ZD7 D23 120  D21 D20 D22 D19 Q12 IC10 4023 1 D18 0.1 Q11 560pF IC9 4013 D17 1 1 1k 2200 Q16 .047 1 0.1 IC6 555 0.1 IC7 555 0.1 VR1 0.1 Q15 10  10k IC3 LM358 .001 IC5 LM358 1 T2 .0047 12k 10k 10k 0.1 390k 10k 1 Q10 100pF D15 0.1 0.1 47k 4.7k 10k 0.1 0.1 1M 1 1M 4.7k RO .001 D16 D4 22k IC1 TL494 K 47k D3 0.1 10uF D7 D6 1.2M 10k 10k D5 100pF D13 560pF 2.2k ZD6 Q9 220k D11 T3 0.1 Q6 10k R1 0.8mm ENCU D1 Q4 0.1 400VDC 1000uF Q3 Q2 Q1 IC8 4017 1 15k 150pF 220pF K S1 A F1 N GPO A LED1 K FRONT PANEL February 1994  31 This interior view of the 200W Inverter highlights the small high frequency transformer & the 100µF high-voltage reservoir capacitor. Note that holes must be drilled in the heatsink flange to clear the mounting screws for the earth lug & Mosfet Q3. (PWM) controller. It contains a sawtooth oscil­lator, two error amplifiers and a pulse width modulation compara­ tor. It also includes a dead time control comparator, a 5V refer­ence and output control options for push-pull or single ended operation. The components at pins 5 and 6 set the operating frequency of the inverter at about 25kHz. This frequency was selected to obtain the maximum power from the transformer. The PWM controller generates variable width pulses at pins 9 and 10 and these are buffered by the triple paralleled buffers of IC2, to drive the gates of Mosfets Q1 and Q2 via 10Ω resistors. 32  Silicon Chip Mosfets Q1 and Q2 drive the primary winding of transformer T1 which has its centre-tap connected to the +12V battery supply. Each Mosfet is driven with a complementary square wave signal so that when Q1 is on, Q2 is off and when Q2 is on, Q1 is off. The resulting waveform on the primary is stepped up by the secondary winding. Zener diodes ZD1 and ZD2 protect Q1 and Q2 from overvol­ tage. They operate at follows: when each Mosfet switches off, the transformer applies a positive voltage transient to the drain. If this exceeds the breakdown voltage of the zener (75V), it con­ducts and turns on the gate of the Mosfet which effectively then clamps the transient. The diodes in series with each zener prevent negative gate voltages. The stepped-up secondary voltage of T1 is rectified by high-speed diodes D3-D6 and filtered by the 100µF 385VDC capaci­tor. Voltage feedback A voltage divider comprising a 1.2MΩ resistor and a 3.3kΩ resistor monitors the high voltage DC from the inverter and drives op amp IC5a. This in turn drives linear optocoupler IC4. This device provides electrical isolation between input and output and drives IC3b, another op amp. Note that IC5b, the second op amp in the LM358 package, is not used. continued on page 37 SILICON CHIP If you are seeing a blank page here, it is more than likely that it contained advertising which is now out of date and the advertiser has requested that the page be removed to prevent misunderstandings. Please feel free to visit the advertiser’s website: www.jaycar.com.au SILICON CHIP If you are seeing a blank page here, it is more than likely that it contained advertising which is now out of date and the advertiser has requested that the page be removed to prevent misunderstandings. Please feel free to visit the advertiser’s website: www.jaycar.com.au SILICON CHIP If you are seeing a blank page here, it is more than likely that it contained advertising which is now out of date and the advertiser has requested that the page be removed to prevent misunderstandings. Please feel free to visit the advertiser’s website: www.jaycar.com.au SILICON CHIP If you are seeing a blank page here, it is more than likely that it contained advertising which is now out of date and the advertiser has requested that the page be removed to prevent misunderstandings. Please feel free to visit the advertiser’s website: www.jaycar.com.au 36  Silicon Chip by transis­tors Q15 and Q16. We shall discuss the 1MHz source later in this article. The secondary output of transformer T2 is rectified using four 1N4148 switching diodes (D20-D23) and filtered with a 0.1µF capacitor. The resulting DC is regulated by 12V zener diode ZD7 and then powers IC5 and part of IC4. Current limiting Fig.6: primary winding details for the 25kHz DC-DC inverter section (T1). The four primary coils are quadrifilar wound with 1.25mm diameter enamelled copper wire. Note that the secondary winding is not shown. Its inputs (pins 2 & 3) are tied to pin 4 on the PC board. Trimpot VR1 is used to adjust the DC error signal from IC4 and thereby sets the high voltage DC rail. The signal from VR1 is amplified by IC3b and applied to the internal error amplifier in IC1 via diode D8 to control the pulse width modulation drive to the Mosfets. If the high DC voltage becomes greater than +340V, the pulse width drive is reduced. Similarly, if the voltage drops below +340V, the pulse width is increased until the correct voltage is achieved. Note that op amp IC5 and the high voltage side of IC4 cannot be powered from the 12V battery since the high voltage circuitry has to be fully floating. Hence they need their own isolated DC supply. This is provided by transformer T2. This transformer is driven at 1MHz via a .0047µF capacitor The current drain of the DC-DC Inverter is kept in check by op amp IC3a. This monitors the voltage drop across the 430µΩ sensing resistor connected between the sources of Q1 and Q2 and the negative supply (ie, 0V). IC3a amplifies the voltage drop across this resistor (which is a set length of specific diameter wire) by 391 so that only a very small voltage need appear across the resistor before overcurrent occurs. IC3a’s output is fed to the pin 16 input of IC1 via diode D7. It effectively overrides the voltage control of IC3b should the current rise above 30 amps. Dead time Dead time mightn’t sound like a good idea but is necessary in pushpull inverters otherwise the transistors or FETs can destroy themselves. This can happen because at the moment of switch-over, both Mosfets can be on. The “dead time” comparator at pin 4 prevents the push-pull outputs at pins 9 and 10 from changing over at the same time. It does this by providing a brief delay between one output going low and the other output going high. The dead time is also increased when power is first applied to achieve a slow start up. Initially, the 10µF capacitor between pins 13 and 14 and pin 4 is discharged. This forces a 100% Fig.7: winding details for the five toroid isolating transform­ers. dead time, with both outputs at pins 9 and 10 off. As the capacitor charges via the 47kΩ resistor to ground, the dead time is reduced slowly until it reaches its minimum value. Under-voltage protection is provided to prevent the battery from being discharged too much. Pin 2 of IC1 monitors the battery voltage via a voltage divider comprising 10kΩ and 12kΩ resistors. When the battery drops to below about 10V, the outputs at pins 9 and 10 switch off to shut down the circuit. H-pack output As discussed previously, four Mos­ fets are connected in an H-configuration across the high voltage supply. Mosfets Q3, Q4, Q5 and Q6 are driven by identical transformer coupled gate driv­ers to provide isolation from the low voltage circuitry. The gate driver for Q3 consists of transformer T3, diodes D9 and D10, transistor Q7, zener diode ZD3 and the 220kΩ resistor and 100pF capacitor. To switch on Q3, we apply a 1MHz signal to the primary side of T3. Its secondary voltage is then rectified by D9 and filtered by the 100pF capacitor. The resulting DC signal is fed via diode D10 to the gate of Q3, while zener diode ZD8 provides gate voltage clamping at 15V. So while the 1MHz signal is applied to T3, Q3 is on. To turn Q3 off, the 1MHz signal to T3 is removed but this does not ensure a sufficiently rapid switch-off. This is where Q7 comes into play. The 100pF capacitor discharges via the 220kΩ resistor until the base of transistor Q7 goes 0.7V below its emitter. Q7 then switches on to quickly discharge the gate ca­pacitance of Q3 and ensure a rapid turn-off. As mentioned in the description of Fig.8: mounting details for the Mosfets. Note that Mosfets Q1 & Q2 are also fitted with a finned heatsink. February 1994  37 PARTS LIST 1 plastic instrument case, 200 x 155 x 65mm 1 aluminium panel, 195 x 63 x 2mm 1 Dynamark front panel label, 195 x 63mm 1 PC board, code 11309931, 171 x 141mm 1 finned heatsink, 55mm long x 105mm wide (Altronics Cat. H-0522 or equivalent) 1 5AG panel mount fuseholder 1 30A, 5AG fuse 1 panel mount SPST rocker switch 1 5mm LED bezel 1 miniature mains power point (Clipsal NO.16N or equivalent) 1 30A red battery clip 1 30A black battery clip 3 cable ties 2 cord-grip grommets for 3.5mm dia. wire 3 ring type crimp lugs (blue, 4mm stud) 4 TO-220 mica washers plus insulating bushes plus screws & nuts 2 TOP-3 mica washers plus insulating bushes plus screws & nuts 2 Philips ETD34 ferrite transformer cores (2 off 4312 020 37202) (T1) 1 Philips ETD34 coil former (1 off 4322 021 33852) 2 Philips ETD34 mounting clips (2 off 4322 021 33892) 5 Philips RCC6.3/3.8/2.5 3F3 ring cores (5 off 4330 030 34971) (T2-T6) 5 3mm dia. machine screws, nuts & star washers 5 6BA nylon screws & nuts Wire & cable 1 1.5m length red heavy duty cable (41 x .32mm, DSE Cat. W-2286 or equivalent) 1 1.5m length black heavy duty cable (41 x .32mm, DSE Cat. W-2288 or equivalent) 1 200mm length blue 10A 240VAC mains wire 1 200mm length brown 10A 240VAC mains wire 1 150mm length red hookup wire 1 150mm length blue hookup wire 1 1m length 1.25mm dia. enamelled copper wire 1 16m length 0.4mm dia. enamelled copper wire 1 300mm length 0.8mm dia. enamelled copper wire 1 500mm length 0.8mm dia. tinned copper wire 1 1m length 0.2mm dia. enamelled copper wire the block diagram, Q3 and Q6 switch on and off together and Q4 and Q5 switch on and off together. Consequently, their respective transformers (T3 and T6 and T4 and T5) are driven together. However, each pair of trans­ formers is connected out of phase on the PC board to provide even loading on the transformer drivers. In order to drive the T3-T6 transformers, we need to produce bursts of 1MHz signal every 10ms but only for 70% of the time; ie, for 7ms. In addition, the bursts need to be directed alter­nately to T3 and T6 for one 10ms period and to T4 and T5 for the second 10ms period. Five ICs produce the requisite 50Hz bursts of 1MHz signal. IC6 is a 7555 timer connected to oscillate at 1kHz and it drives IC8, a 4017 decade counter with 10 decoded outputs. The 5, 38  Silicon Chip Semiconductors 1 TL494 switchmode controller (IC1) 1 4050 CMOS hex buffer (IC2) 2 LM358 dual op amps (IC3,IC5) 1 IL300 linear optocoupler (IC4) 2 7555 CMOS timers (IC6,IC7) 1 4017 CMOS decade counter decoder (IC8) 1 4023 CMOS dual D-flipflop (IC9) 1 4023 CMOS triple 3-input NAND gate (IC10) 2 BUK436-100A N-Channel Mosfets (Q1,Q2) Philips 4 BUK655-500B N-Channel FRED FETs (Q3-Q6) Philips 4 BC557 NPN transistors (Q7-Q10) 3 BC338 NPN transistors (Q11,Q13,Q15) 3 BC328 PNP transistors (Q12,Q14,Q16) 19 1N4148, 1N914 switching diodes (D1,D2,D7-D23) 4 BYW95C 600V 3A fast diodes (D3-D6) Philips 2 75V 400mW zener diodes (ZD1,ZD2) 4 15V 400mW zener diodes (ZD3-ZD6) 1 12V 400mW zener diode (ZD7) 1 5mm red LED (LED1) Capacitors 1 2200µF 16VW PC electrolytic 1 1000µF 25VW PC electrolytic 1 100µF 385VDC electrolytic (Philips 2222 052 58101) 2 10µF 16VW PC electrolytic 2 2.2µF 100V MKT polyester 14 0.1µF MKT polyester 1 0.0047µF MKT polyester 2 0.001µF MKT polyester 4 560pF MKT polyester 1 220pF ceramic 1 150pF ceramic 4 100pF ceramic Resistors (0.25W 1%) 1 1.2MΩ Philips VR37 3 1MΩ 7 10kΩ 1 390kΩ 2 4.7kΩ 4 220kΩ 1 3.3kΩ 1 150kΩ 1 2.2kΩ 1 56kΩ 2 1kΩ 2 47kΩ 1 820Ω 1 22kΩ 1 120Ω 1 15kΩ 3 10Ω 1 12kΩ Miscellaneous Insulating tape, heatsink compound 6 and 7 counts of IC8 are ORed with diodes D17, D18 and D19, so that the input pins to NAND gate IC10a are high whenever pins 1, 5 or 6 of IC8 are high. These three outputs are high for three counts in 10 or for 30% of the time. Consequently, after inver­sion by gate IC10a, the output is high for 70% of the time, which is what we want. IC10a drives pins 8 and 11 of gates IC10b and IC10c. Pins 1 and 13 of IC10b and IC10c respectively connect to the complemen­tary outputs of IC9, a 4013 D-flipflop. This flipflop toggles its Q and Q-bar outputs each time it receives a clock pulse from pin 5 of IC8. This occurs every 10ms. The remaining inputs of IC10b and IC10c connect to a 1MHz oscillator, IC7, another 7555 timer. IC10b and IC10c can only pass the 1MHz signal when their other two inputs are both high. This occurs 70% of the time for each alternate 10ms period. For example, the output of IC10b passes the 1MHz signal during one 10ms period and the IC10c output passes the 1MHz signal for the second 10ms period. The output (pin 9) of IC10b is buffered by complementary transistors Q11 and Q12 to drive T4 and T5 via separate 560pF capacitors. Similarly, the pin 10 output of IC10c is buffered by Q13 and Q14 to drive T3 and T6 via separate 560pF capacitors. Let’s now recap on the circuit operation. Mosfets Q1 and Q2 are driven by IC1 at 25kHz to step up the 12V to 340V DC which is regulated and otherwise current limited. Then the H-pack Mosfets are switched to provide a 50Hz modified square wave with an output close to 240VAC RMS. Power for the circuit is obtained from the 12V battery via a 30-amp fuse which supplies the inverter transformer T1 direct­ly. The low current part of the circuit is then supplied via switch S1 and a 10Ω decoupling resistor. A 2200µF capacitor across the supply ensures that the heavy switching currents to the DC-DC converter do not produce voltage fluctuations. A LED connected across the supply in series with a 2.2kΩ resistor indicates when power is on. Construction The 200W Inverter is housed in a plastic instrument case measuring 200 x 155 x 65mm. Most of the circuit components are mounted on a PC board which measures 171 x 141mm (code 11309931) – see Fig.5. Construction of the inverter involves winding several coils and a transformer, assembling the PC board and a small amount of hole drilling and wiring. Note: we do not recommend this project to inexperienced kit builders. Construction can begin by checking the PC board against the published pattern. Look for any broken tracks or shorts and repair any faults now to avoid problems with the circuit opera­ tion later on. Note that 3mm holes should be drilled for the battery supply connections adjacent to transformer T1. If these are not drilled, drill them now. Solder a 3mm brass nut underneath each of these holes, on the copper side of the board. The PC stakes and links can now be installed. Note that there are three types of links and it is important to install them in the correct positions. 0.8mm enamelled copper wire is used for the high voltage sections of the circuit to help provide greater safety since they present less chance of accidental contact when the circuit is running. Most enamelled copper wire is selfflux­ing, meaning that the enamel will strip under heat from a solder­ing iron. However, make sure that each solder joint is a good one. Now all the ICs, resistors and diodes can be inserted. Note that resistor R0 should not be installed at this stage, as it may not be required. More about this point later. Be careful with the orientation of the ICs and diodes and be sure to insert the correct type of zener diode in each position. Now insert the transistors, noting that there are three different types, so be careful to place them in the correct positions. Insert all the capacitors, taking care with the orien­tation of the electrolytics. This waveform shows the 7.5ms bursts of 1MHz signal from pins 9 & 10 of IC10. These signals are fed to the toroid isolating trans­formers, rectified & used to turn on the H-pack Mosfets. This oscilloscope photo shows the gate drive signals to Mosfets Q1 & Q2. Top trace is gate of Q1; lower trace, gate of Q2. Note the time interval between the respective gate pulses to Q2 & Q2, to ensure “dead time”. Winding the coils Transformer T1 is wound using 1.25mm diameter enamelled copper wire. Fig.6 shows how it is done. Locate pins 1, 2, 3 and 4 of the transformer bobbin and terminate four wire ends to these pins. Wind the four wires together (ie, quadrifilar winding) and make three turns. Terminate the wire ends at pins 14, 13, 12 and 11. Insulate the winding with a layer of paper and a layer of insulating tape. Now the secondary is wound on with 0.4mm enamelled copper wire. Terminate one end of the wire to pin 7 and wind on 115 turns neatly, side by side. Insulate between each layer with insulating tape before winding the next layer and make sure that each layer is wound in the same direction as the last. Finally terminate the wire end on pin 8. That completes the secondary winding. The transformer is assembled by in- This is the 240VAC output waveform from the inverter when driv­ing a 160 watt lamp load. Note that this wave shape changes very little, regardless of the load. serting the ferrite cores into each end of the bobbin and fitting the clips at the ends to hold them in place. Check that the faces of the ferrite cores are absolutely clean before assembling them. Toroids T2 and T3-T6 are each wound using 0.2mm enamelled copper wire, as shown in Fig.7. Each February 1994  39 RESISTOR COLOUR CODES ❏ ❏ ❏ ❏ ❏ ❏ ❏ ❏ ❏ ❏ ❏ ❏ ❏ ❏ ❏ ❏ ❏ ❏ No. 3 1 4 1 1 2 1 1 1 7 2 1 1 2 1 1 3 Value 1MΩ 390kΩ 220kΩ 150kΩ 56kΩ 47kΩ 22kΩ 15kΩ 12kΩ 10kΩ 4.7kΩ 3.3kΩ 2.2kΩ 1kΩ 820Ω 120Ω 10Ω winding is wound tightly with the wires as close together as possible. Keep the two windings separate to ensure electrical isolation between them. T3, T4, T5 and T6 must be wound identically. Final PC board assembly Transformer T1 and the toroid coils can now be installed. When inserting T1, make sure that it is oriented correctly. The 1.25mm diameter primary winding end must be adjacent to Mosfets Q1 and Q2. The toroids are secured with Nylon screws and nuts. Do not use metal screws since they will reduce the isolation between the primary and secondary windings. Be sure to orient the toroids correctly on the PC board; ie, the 12-turn secondaries should be adjacent to the associated 220kΩ resistors. Mosfets Q1-Q6 can now be inserted into the PC board and sol­dered. The lead length for each Mosfet should be 10mm. Position the PC board in the case and check the four inte­gral standoffs used to support the PC board in place. Use a large drill to shorten all the unused standoffs so that the PC board will sit neatly in position. Secure the PC board in place with self-tapping screws and slide the rear metal panel into its slot. Mark out the positions for the Mosfet mounting 40  Silicon Chip 4-Band Code (1%) brown black green brown orange white yellow brown red red yellow brown brown green yellow brown green blue orange brown yellow violet orange brown red red orange brown brown green orange brown brown red orange brown brown black orange brown yellow violet red brown orange orange red brown red red red brown brown black red brown grey red brown brown brown red brown brown brown black black brown holes on the rear panel. Drill these holes to suit the 3mm mounting screws. While you’re at it, drill and file the two cord grip grommets and the earth lug (3mm). The finned heatsink is also retained with four screws and nuts, two at the top and two at the bottom edge. Drill the necessary holes in both the rear panel and heatsink and the holes in the heatsink for Mosfets Q1 and Q2. The heatsink fins will also need drilling out with holes large enough for the screw heads for Mosfet Q3 and the earth lug. Remove any burrs around the holes, particularly where the Mosfets mount, to prevent punch-through of the mica insulating washers. You will need to secure the earth terminal screw and the screw for Q3 with nuts before attaching the heatsink to the rear panel. This is because these screws cannot be inserted once the heatsink is on. Apply a smear of heatsink compound between the mating faces of the heatsink and rear panel to ensure good heat transfer. Fig.8 shows the mounting details for each of the Mosfets (Q1-Q6). They need to be isolated from the panel with a mica washer and insulating bush. When you have tightened down the screw and nut, set your multi­meter on a high “Ohms” range and check that the metal tab of each device is indeed isolated from the rear panel and heatsink. 5-Band Code (1%) brown black black yellow brown orange white black orange brown red red black orange brown brown green black orange brown green blue black red brown yellow violet black red brown red red black red brown brown green black red brown brown red black red brown brown black black red brown yellow violet black brown brown orange orange black brown brown red red black brown brown brown black black brown brown grey red black black brown brown red black black brown brown black black gold brown Work can now be done on the front panel. Use the front panel label as a guide to positioning the 30-amp fuse holder, switch, LED bezel and mains socket. Drill out the holes for each of these, then affix the label and cut out the holes with a reamer and sharp knife. Secure the fuse holder, switch, LED and LED bezel and the mains socket to the front panel, ready for wiring. Follow the wiring diagram carefully and use the correct wire, as specified. If the two cordgrip grommets do not grip the wires secure­ly, use some heatshrink tubing to increase the wire diameter. Do not use one grommet to secure both wires since there is a pos­ sibility that the wires may short out. The heavy duty hook-up wires (41 x 32mm) from the negative terminal of the battery and the fuseholder are fitted with crimped lugs and then secured with screws to the PC board (these screws go into the nuts previously soldered to the underside of the board). Use cable ties to tidy up the wiring when completed. Fit the battery leads with 30A battery clips. Testing Warning! Exercise extreme caution when doing measurements on this inverter. The voltages can be lethal. Use only one hand and do not touch any part of the circuit, particularly if Fig.9: actual size artwork for the PC board (code 11309931). Check your etched board for defects by comparing it against this pattern & correct any defects before installing the parts. you have connected an oscilloscope earth lead. Always check the voltage between TP1 and TP2 and wait until the voltage dies to a safe level (less than 30V) before touching any part of the circuit. Before applying power, check your work carefully and verify that your wiring and parts layout is the same as the wiring diagram of Fig.5. For the initial tests, it is best to use a 12VDC power sup­ply. Connect the +12V to switch S1, on the same side that LED 1 connects (ie, we don’t want power applied to T1 or to Mosfets Q1 & Q2). With switch S1 off, apply power. Check that +11.4V is present at the supply pins of all the ICs; ie, pins 8,11 &12 of IC1, pin 1 of IC2, pin 8 of IC3 and IC5, pin 6 of IC4, pins 4 & 8 of IC6 and IC7, pin 16 of IC8 and pin 14 of IC9 and IC10. There should also be 12V across ZD7. A DC measurement across ZD3, ZD4, ZD5 and ZD6 should show about 5.4V. Similarly, between ground and the gate of Q1 and ground and the gate of Q2 should show about 5.0V. If you have an oscilloscope, the wave­forms in the accompanying oscilloscope photographs should be compared. If all these tests check out OK, you are ready for a high voltage test. Disconnect the 12V supply used for initial testing although, if it can deliver 8 amps or more, it can be used for the initial high voltage tests too. Rotate trimpot VR1 fully anticlockwise. This will set the high voltage to a minimum. Place the lid on the inverter and connect it to your 12V supply or 12V battery. Switch on S1 brief­l y and then turn off. The reason for having the lid on the in­verter at initial switch-on is that if there is something wrong with the high voltage side of the circuit, one or more components may blow. So the lid on the inverter will protect your eyes! Alternatively, wear eye protection goggles. Now take off the lid and remember that the circuit is now dangerous. Check the DC voltage between test points TP1 and TP2. It should be above 100V DC but falling. Do not touch any part of the circuit until the voltage drops to a safe level (below 30V). Now apply power again and check the voltage between TP1 and TP2. Watch the meter and adjust VR1 slowly until the voltage is set at 340V DC. You can now install the lid and load test the unit. Check that it will drive 240VAC light bulb loads up to 200W. If the fuse blows when powering a 200W load, the R0 (1MΩ) resistor should be installed to slightly increase the dead SC time for IC1. February 1994  41 Electronic Engine Management Pt.5: Oxygen Sensors – by Julian Edgar A major incentive for adopting engine management systems was to meet the strict exhaust gas emissions legislation enacted in several geographical areas – notably the huge Californian market. To meet these strict emissions levels, manufacturers had to start tuning their cars to meet these criteria, rather than optimising for power and economy. The initial response by manufacturers to Australian legislation was often half-hearted, with Australian Design Rule (ADR) 27A back in the mid-1970s giving us cars which drank fuel with a voracious thirst, overheated and stalled in traffic. This reflected poor design adaptation rather than any intrinsic prob­lems with the new regulations. Unleaded petrol The coming of unleaded petrol (ULP) in 1986 meant that engines had to be redesigned to run on lower octane fuel which lacked lead. For some local makers, their old engines simply couldn’t be updated and so new engines were introduced. Holden replaced its venerable red/blue/black 202 (3.3 litre) engine with the Nissan 3.0 litre straight six, for example, before switching to an American-designed 3.8-litre V6. As well using the new fuel, the car manufacturers also had to use a catalytic converter. A catalytic converter changes the “colour” of several of the more noxious pollutants to “green”, thereby benefiting the environment. However, leaded fuel will poison a catalytic converter and so must not be used. (Inciden­tally, ULP will always give a black tailpipe – irrespective of mixture strength). Air-fuel ratio This engine uses a single-wire (unheated) oxygen sensor. It is shown bolted through the top of exhaust manifold, just to the right of the turbocharger assembly. 42  Silicon Chip Also required for efficient catalytic converter operation is an air-fuel ratio that’s very close to stoichiometric (14:1). This means that, for the catalytic converter to work best, 14kg of air (or 10,000 litres) must be mixed with every litre of petrol. Incidentally, the stoichiometric ratio – where theoreti­ cally best combustion occurs – varies from 14:1 to 14.7:1, ac­cording to the reference used! The authoritative Bosch Automotive Handbook lists it as 14:1. Fig.1 shows the relationship between varying air/fuel ratios around stoichiometric and the production of the pollutants carbon monoxide, hydrocarbons and oxides of nitrogen. An example of a heated oxygen sensor from a Nissan engine. Note that there are three leads running back to the plug connector. The stoichiometric ratio isn’t, however, the best for either maximum power or economy, with the mixture needing to be richer or leaner respectively to achieve this. Mixture feedback loop Car manufacturers were therefore faced with a dilemma – did they design for power, economy or emissions? They solved this by using a feedback loop which allowed them to have their cake and eat it too. At constant throttle settings (that is, cruise), the exhaust gas is monitored for mixture strength and information from the sensor fed back to the ECM which in turn controls injec­tor pulse width openings to give a stoichiometric mixture. Fig.2 shows the structure of the feedback loop. At full throttle (sensed by the throttle position switch), the system goes open loop, with the exhaust gas oxygen (EGO) sensor ignored and the mixture suitably enriched for power. Conversely, lean mix­tures are used during a trailing throttle. The EGO sensor keeps track of all Fig.1: the relationship between air/fuel ratio & the production of various pollutants. February 1994  43 CONTROL UNIT FEEDBACK SIGNAL INJECTION PULSE of that point. Fig.5 shows the voltage response of a typical EGO sensor. Note that its output voltage does not directly follow oxygen concentration, especially for lean mixtures. OXYGEN SENSOR Mixtures revealed OXYGEN SENSOR INJECTOR FUEL INJECTION COMBUSTION ENGINE Fig.2: the EGO sensor feedback loop. At full throttle, the system goes open loop & the EGO sensor is ignored. of these mixture varia­tions. It can be one of two types – titanium or zirconia oxide. The zirconia type is more frequently used and generates a voltage output. A cross-sectional view of a typical zirconia EGO probe is shown in Fig.3. Its operating temperature is from 300°C upwards and it is sometimes electrically heated to bring it up to this temperature. Its performance in unheated mode is usually satis­factory, though, and so some manufacturers run it like this. The other type of EGO sensor – the titanium probe – must always be electrically heated. Instead of generating its own voltage output, the titanium probe changes its resistance in response to different oxygen levels in the exhaust. It is mounted close to the engine in the exhaust manifold to ensure that it is quickly heated to operating temperature – see Fig.4. Both probe types are calibrated so that their output chang­es rapidly around the stoichiometric point and is symmetrical in response to either side BUSHING (ELECTRODE) The most interesting aspect of EGO sensors is that it is easy to access their output and then see for yourself the mixture variations that occur as the car is driven. It’s a bit like gaining sight after being blind – suddenly you can see the cold-start and full throttle enrichment cycle working, the overrun injector cutoff, the time when the computer is in closed loop mode, and when the computer goes open-loop. And in a car running modified EFI – whether by chip rewrit­ing or cruder means – it can be clearly seen where rich or lean points occur in real driving conditions. Obtaining a readout from the common zirconia EGO probe is easy, because the commonly-available LM­ 3914 LED display driver IC seems almost custom designed for the purpose. By following the attached circuit, a 10-LED display mixture meter can be easily and cheaply constructed – see Fig.6. The voltage output from the EGO sensor is usually between 0-1V, with the sensor in most cars giving 0.5V at the stoichiometric point. The IC uses an internal reference of 1.25V and this is easily reduced to 1.0V by a trimpot (VR1). Inside the LM3914 is a series of op amp comparators and these each compare the signal voltage from the EGO with a divided reference signal. Each op amp in turn drives an LED (LEDs 1-10) and this produces a moving TERMINAL SUPPORT (LEAD WIRE INSULATION) LEAD WIRE ATMOSPHERE SPRING EXHAUST MANIFOLD Fig.3 cross-sectional view of a typical EGO sensor. 44  Silicon Chip Fig.4: the EGO sensor is bolted into the exhaust manifold, close to the engine. O2 SENSOR VOLTAGE Obtaining a readout EXHAUST GAS ZIRCONIA PIPE EXHAUST MANIFOLD CO CONCENTRATION O2 CONCENTRATION RICH THEORETICAL AIR/FUEL RATIO LEAN Fig.5: the output from a typical EGO sensor in response to O2 levels. display as the input voltage rises or falls. Pin 9 controls the display mode. Leaving pin 9 open cir­cuit produces a dot display, while tying pin 9 to pin 3 produces a bargraph display. The 680Ω resistor sets the display brightness. The components can be bought individually and mounted on a board, or the Jaycar Car Battery Monitor kit (which uses the same IC and comes with 10 square LEDs) can be modified to work in the 0-1V range. The circuit shown is about the simplest possible. Variations include using diodes to limit voltage spikes and slowing the response time of the meter by using capacitors to filter the input signal. Connecting the meter to the sensor is straightfor­ ward – just connect it in parallel with the ECM. If the EGO sensor is a 3-wire type, then use the workshop manual (or a high input-impedance multimeter) to sort out which is the sensor output wire. If the EGO sensor is a variable- LED1-10      10 11 12 12V  13   14 15 16   17 18 1 3 INPUT FROM OXYGEN SENSOR VR1 5k 5 IC1 LM3914 6 7 2 4 8 680  Fig.6: the readout for the oxygen sensor is based on IC1, an LM3914 dot/ bar display driver IC. It functions as a simple LED voltmeter. resistance type (rare), then obviously the LED meter will be inappropriate in this form. Finally, connect 12V and earth and the meter should come alive when the sensor is up to temperature The meter’s output display will depend on the type of ECM your car uses. In closed-loop mode (with the EGO sensor having an input into injector pulse width decisions), the mixture will cycle rich-lean-rich-lean, either at a few Hertz or almost in­stantly back and forth. Alternatively, some cars will cycle for a few seconds and then settle at the “correct” mixture, holding it at the point until a throttle change. Others will require perhaps 60 seconds of constant-speed cruising before holding the mixture steady on the display. However, flooring the right foot will instantly give a rich readout, as the ECM software commands for maximum power are invoked. If your car runs plain ol’ carbies, you can still use an oxygen sensor. It will help if you use ULP in your car, as other­wise the sensor will be prone to lead fouling. Oxygen sensors are quite expensive when new but a car wrecker importing engines directly from Japan will have used sensors available. I bought two sensors in this way for $15 for the pair. Depending on the design of the sensor, either a nut or mounting plate will need to be welded to the exhaust to allow it to be fitted. Place the sensor as close to the engine as possible, making sure that it will get the gas flow from SC all cylinders. The completed mixture display meter. It connects directly to the EGO sensor. February 1994  45 Build the CHAMP: a handy audio amplifier based on a single IC What’s the same size as a 9V battery, more useful than a deck of cards and uses only a half a dozen components? The CHAMP – a Cheap & Handy AMPlifier that will deliver 0.5W into eight ohms from a 9V supply for those little audio projects. By DARREN YATES Well, this is about as small as you can get with standard sized components – a single channel audio power amplifier than will produce 0.5W into eight ohms with a 9V battery and with variable internal gain from 20 to 200. It can also drive a 4Ω loudspeaker at lower power levels and with increased distortion (see Figs.2 & 3). We don’t claim this to be an original design but it is tiny! You’ll be surprised by the number of projects that use an audio amplifier of some kind. Most 46  Silicon Chip of the time, they are only low power modules hanging off the end of some noise-maker but it seems a pity to have to re-engineer the wheel every time an amplifier is needed. This module uses the well-known LM386 audio amplifi­er IC. It’s small and most importantly, cheap. It will fit into the tightest of spaces – you could even glue it to the back of a 9V battery if you wanted to! Mind you, running it from a small 9V battery would not be an economical prop- osition – better to run it from a 6V lantern battery, four 1.5V AA cells or a 9V DC plugpack. Circuit diagram The circuit diagram for the CHAMP is shown in Fig.1. As you can see, there’s not much to it. The power supply can be anything from 4 to 12VDC and connects straight to pins 6 and 4. The input signal is fed to a 10kΩ trimpot and then straight to pin 3 of the IC. The 10µF capacitor connected to pin 7 provides supply decoupling and reduces the effect of hum on the power supply if it comes from a 9V DC plugpack. The gain of the amplifier can be changed from 20 to 200 by changing the value of the 1kΩ resistor at pin 1. Reducing the resistor increases the gain. As presented, the circuit gain is 41 or 32dB. By replacing the resistor with a wire link, the gain becomes 200. A 220µF capacitor couples the out- Fig.1: the circuit is based on IC1, an LM386 audio amplifier IC. The gain of the amplifier is controlled by the 1kΩ resistor on pin 1. +4-12V 10 16VW INPUT 6 VOLUME VR1 10k LOG 2 1k 8 3 1 IC1 LM386 4 100 16VW 7 220 16VW 5 0.1 8W 10 16VW 10  PARTS LIST 1 PC board, code 01102941, 46 x 26mm 6 PC pins 1 10kΩ trimpot Semiconductors 1 LM386 low-power audio amplifier IC (IC1) Capacitors 1 220µF 16VW electrolytic 2 10µF 16VW electrolytics 1 0.1µF 63VW MKT polyester THE "CHAMP" Resistors (0.25W, 1%) 1 1kΩ 1 10Ω Miscellaneous Tinned copper wire, speaker cable, solder. Fig.2: device dissipation vs output power for a 4Ω load. The three curves shown are for 6V, 9V & 12V supply rails. Note that the maximum power output into a 4Ω load is about 0.3mW at 3% THD. Fig.3: device dissipation vs output power for an 8Ω load. In this case, four curves are shown, corresponding to supply voltages of 6V, 9V, 12V & 16V. The maximum power output is about 0.7W at 3% THD. put signal from pin 5 to an 8Ω speaker. You can also use a 4Ω ohm speaker if the supply voltage is 6V or less. A Zobel network consisting of a 0.1µF capacitor and 10Ω 0.25W resistor prevent high-frequency oscilla­tion from occurring due to long speaker leads. Finally, a 100µF 16VW capacitor provides supply decoupling and aids in the operation from a battery supply. Power output will vary depending on the supply voltage and whether a 4Ω or 8Ω loudspeaker is used. The graphs of Fig.2 and Fig.3 show what can be expected with 4Ω and 8Ω speakers at vari­ous supply voltages. 1k VR1 1 SPEAKER + 10uF 10  100uF 220uF 0.1 GND We designed a teensy weensy little PC board for this pro­ject but although it’s small, it’s a snack to put together. The board measures just 46 x 26mm and is coded 01102941. Apart from the LM386 IC, it has two resistors, a trimpot and five capacitors on the board. +4-12V 10uF IC1 LM386 INPUT Construction SPEAKER GND Fig.4 (left): take care to ensure that the IC & electrolytic capacitors are all installed the right way around during the PC board assembly. The power supply to the board can be anywhere in the range from 4-12V DC. Fig.5 at right is a full-size etching pattern for the PC board. Check the board carefully for any defects in the copper pattern such as shorted or broken tracks. If there are any, fix them before proceeding further. Begin the assembly by installing six PC pins at the external wiring points, followed by the two resistors, then the capacitors, trimpot VR1 and the IC. Be sure to install all polarised parts the right way around – ie, the IC and electrolytic capacitors. Testing Connect a 9V power supply to the amplifier module, with your multi­ meter (switch­ed to the 200mA range) in series with one of the leads. Do not connect a loudspeaker at this stage. With no input signal, you should get a quiescent current of about 8-10mA. Any more than this and you should switch off immediately and check the PC board against the overlay diagram to see if you have made any errors. Once everything appears to be OK, connect a loudspeaker and do a “blurt” test. You do this by winding the trimpot anti-clockwise and then putting your finger on the input. This injects a hum/hash signal into the amplifier which is heard as a “blurt” from the speaker. If it blurts, it’s working. Finally, the installation of the CHAMP is basically left up to you. Make sure you keep it away from mains transformers or anywhere large SC amounts of hum are present. February 1994  47 SILICON CHIP If you are seeing a blank page here, it is more than likely that it contained advertising which is now out of date and the advertiser has requested that the page be removed to prevent misunderstandings. Please feel free to visit the advertiser’s website: www.altronics.com.au SILICON CHIP If you are seeing a blank page here, it is more than likely that it contained advertising which is now out of date and the advertiser has requested that the page be removed to prevent misunderstandings. Please feel free to visit the advertiser’s website: www.altronics.com.au SERVICEMAN'S LOG If only the fault would show One of the truisms of service work is that you shouldn’t try to fix a fault that you can’t see. But we are forced to try sometimes, even if we don’t often win. My main story this month tells of the frustrations of trying to work this way. In order to set the scene for this story, I must reintroduce Murphy. Remember Murphy? He’s the pesky little leprechaun who sneaks around, shifting component values up and down, opening and closing dry joints, and generally contributing to erratic be­haviour in electronic devices. He’s not been around lately and I had hoped he had met with some calamity; like digging a hole for another pot of gold, digging it too deep, and pulling it in on himself. No such luck. More likely, he had been away at some lepre­chaun workshop, learning even more devious ways to create havoc. At least that’s how it appeared when he turned up on this partic­ular job. Not only did he waste a lot of time but I was left with a situation whereby the fault was cor- 50  Silicon Chip rected without being sure why things happened as they did. A simple beginning It all started out simply enough. It was a 68cm NEC colour set (model FS-6831S) that was only a few months old and still under warranty. The customer had purchased it over the counter, taken it home and connected it up himself, which is pretty much par for the course these days. And it had performed perfectly from the start. But now there was a problem, although it was straightfor­ward enough; it wouldn’t switch on. The only sign of life was the standby light but operating the “ON” button on either the set itself or the remote control unit had no effect. The failure of both controls was, in itself, not surpris­ ing, since both employ the same mechanism within the set. But it did rule out any fault in the remote control unit or the panel control. More than that I was not prepared to speculate on until I had my hands on the set. The size of the set could have been a problem. I try to avoid house calls as much as possible but transporting a 68cm set can be difficult for some customers and I have to make excep­ tions. Fortunately, this customer was cooperative and was both willing and able to bring the set to the workshop. So that solved that problem. This set is fitted with a fairly routine chassis and its on-off switching system is similar to that used on a number of sets. It employs relay contacts in the mains active and neutral lines, the relays being activated by the central processing unit (CPU). Most designs use a single relay with two sets of contacts but in this set, for some reason which escapes me, they have elected to use two separate relays – RL651 and RL652. Not that this matters a great deal; the relay coils are connected in parallel and driven from the collector of transistor Q651, a 2SC2002. And the base of Q651 goes to pin AA17 of CPU PW8. So, assuming an appropriate voltage appears at pin AA17, Q651 should turn on, activate the relays, and close the two mains leads. Finding the fault was no big deal. I checked whether the “ON” control produced a voltage on the base of Q651 and, yes, it did. So why wouldn’t the relays operate? Quite simple really; Q651 was open circuit. I didn’t have a 2SC2002 in stock but a look the specifications suggested that a BC639, which has somewhat higher ratings, should do the job. And it did. So the customer duly collected his set and went on his way. And that seemed like the end of the story. Which it was, for the next couple of months. Then the customer was back on the phone, explaining (almost apologetically) that he was having trouble playing his VCR through the set. He was unable to get a recognisable picture; just a mess of streaks and patches of colour, with only an occasional hint of a locked image. But he was quick to add that he didn’t think this was anything to do with the previous fault. Oh, for more like him! Tuning problem? And so began what was to be a long and tedious search for this new fault. My first query was whether the tuner programming had been upset in any way, such that channel 1, used by this VCR, was off frequency. As with most modern sets, this one has a search function which looks for each signal In short, I could find nothing wrong and the customer col­lected the set and took it home. But we had achieved nothing and I wasn’t really surprised when he was on the phone again with the same tale of woe. But he was a gluten for punishment. “Suppose I bring both the set and the VCR up to the shop?” Well, he’d forestalled my thinking there, except that I was envisaging having to make a house call. One way or another, I had to see the two units working together in order to see the actual fault, which Murphy had contrived to hide from me so far. Both units together in turn, adjusts the tuner, and allocates the channel to a selected button. If it was simply a tuning problem, it should be possible for the customer to correct the condition himself and avoid the expense of a service call. I therefore advised him on how to check this – simply run a pre-recorded tape in the VCR and ini­ tiate the search routine. The system would eventually detect the carrier from the VCR and treat it as any other RF signal. I left it with him but he was on the phone a couple of days later, to report no success. The system had apparently gone through its routine correctly but could not correct the signal from the VCR. After a few more probing questions, I began to feel that there must be a fault in the VCR. And so I suggested he drop this in for a check. This he did and at the first opportunity I connected it to one of my monitors, pushed in a test tape and set things in motion. Result: a prefect picture. I put it through all its paces and let it run for several hours after which it still gave a perfect picture. So that ruled that out. When the customer called to collect it, I demonstrated it to him, then went through the tuner programming routine with him again. It was the only explanation I could think of and I sug­gested he give it another try. But again, no joy. He was on the phone the next day com­plaining that everything was exactly the same. He then suggested that he should bring the set into the shop and, since it was his idea, I readily agreed. And so the set duly arrived but without the VCR. This seemed quite logical; I was convinced that the VCR was OK, which meant that it had to be the set. I set the colour bar generator for channel 1, fed it in, and put the set through its search and program routine. And it went exactly as it should. Granted, channel 1 was slightly out but that didn’t really surprise me. The colour bar generator is generally more accurate than the average VCR, the channel fre­quency generated by some being best described as “nominal”. Not that it really matters, as long the set is accurately tuned to whatever the VCR is delivering. And so I finally finished up with both units on the bench, with the customer standing by while I set things up. I put it through the search routine and, as I expected, the VCR output was slightly off-tune for channel 1 although that was easily allowed for. And so, at long last, I should now at least know what the problem looked like. But not a bit of it. Murphy saw to that. Would you believe that the whole setup worked perfectly? Because that is what happened and I was just as confused as the customer. So all I could suggest was that he take it all back home and try again. If I had inadvertently done something to cure the fault, then well and good. If not – well, I had a pretty good idea what the next step would have to be. No prizes for guessing. The customer was on the phone the next day and the problem was just the same as before. At that stage, I could only speculate that it was something peculiar to the house setup, or some weird local interference. In any case, it left no alternative; I had to visit the house, see the problem for myself, and take it from there. So an appointment was made and I faced up to the problem in the customer’s lounge room. He turned the system on and pushed in a tape. The result was pretty well as he had described it but the symptom which struck me most forcibly was that, on the few occasions when the picture tried to lock, it was pulling very badly. My impression was that either the RF out of the VCR was hopelessly unstable or that the TV set was being overloaded in some way. My first step was to put the set through the search program with the February 1994  51 antenna disconnected, so that all it had to search for was the VCR output. Well, it went through the motions but didn’t want to look at the VCR signal, simply skipping over it and going round again. I was still trying to make sense of all the symptoms, when the customer happened to mention that the lead between the VCR and the TV set was not the original (black) one supplied with the VCR. Apparently, the original had been temporarily mislaid at some time and this was a white “el cheapo” one from the local electronics store. Could it be the culprit? I didn’t think so but was prepared to clutch at any straw. I took a closer look and found that moving the cable near the socket on the TV set could cure the fault. In the meantime, the owner had fished out the original cable, so I substituted it. When I did, the system immediately came good. So what would you think? Faulty white lead. Of course. Except 52  Silicon Chip that I could find nothing wrong with it; both the active and braid circuits were continuous and appeared to be reliable. I then tried moving the black lead at the TV antenna socket and, wouldn’t you know it, the fault was back. Faulty socket? Yes, it was. The active (female) contact had spread and was not mating reliably with the male contact. But this fault was not working as one might reasonably imagine from the description. The poor contact in the socket was not causing the fault – it was curing it. An overload problem Remember my impression that the system was being overload­ed? Well, that idea was suddenly starting to make sense. I tight­ ened the sloppy socket contact and this restored the fault in all its glory, regardless as to whether the black lead or the white lead was used. I was beginning to sense victory now. From my kit I fished out a 20dB attenuator, one of several values which can be very useful in some sticky situations. I inserted this between the VCR and the TV set and bingo, we had a perfect picture. OK, so we had an overload problem. But why? My immediate reaction was a fault in the AGC circuit of the TV set. And the first thing to investigate was the preset AGC pot. It could be faulty (intermittent?) or it could be wrongly set, although this latter theory seemed unlikely. I removed the attenuator, then pulled the back off the set and tracked down the AGC pot. And one glance was enough to raise my eyebrows. One normally finds these pots set at around mid-travel but this one was almost fully clockwise. It is not often that this pot needs adjustment, as the factory setting should cope with 99.9% of conditions. But if it does need to be reset, the normal procedure is to first turn it fully anticlockwise, which produces maximum AGC voltage and a snowy picture, even on strong signals. The pot is then advanced until, usually quite suddenly, the snow vanishes and there is a clean picture. And normally, this setting will hold for a wide range of signal strengths. Which was what I did, using the signal from the VCR. And everything went according to the book, including a near mid-setting for a clean picture. I reconnected the antenna and re­ peated the check with off-air signals. Again, everything went according to Hoyle. So that was the answer; an incorrectly set AGC pot – a mistake which almost certainly occurred in the factory during final testing. It was all very gratifying except that I have no explana­tion for all the variations of behaviour. Why did the set work perfectly for the first eight or nine months of its life? Why did it work on my bench and not in the customer’s home? And so on. I’ve tried to work out the answers but I’m afraid they elude me. All I can do is take the easy way out and blame Murphy. The TV that flipped My next story is relatively simple. It did not involve any great detective work to solve but it was unusual. It involved a National colour set, model TC-2178, fitted with an M14 chassis, and about six or seven years old. The customer – the lady of the house in this case – rang to complain that “the set was doing funny things”. Not being quite sure what she meant, my imagination ran wild for few moments but I eventually pinned it down in more precise terms. This set uses a fairly simple channel selection system – two UP/DOWN pushbuttons on the front of the set which can select any one of 12 pretuned channels. There is also a remote control unit with a similar UP/DOWN facility plus a set of 12 buttons (one for each channel) and the usual volume and on/off controls. The problem was that the set had developed a penchant for position 11. It didn’t matter which position was initially se­lected; as soon as the UP or DOWN button was released, the set would immediately move to position 11, which was blank (only six positions were active – five for off-air signals and one for the VCR). In fact, the only way they could hold the set on a particu­lar channel was to select it via the remote control unit and then hold that button down; something which became a mite tiresome after a couple of hours. Well, it was new one on me and I could think of no explana­tion off the top of my head. So I could only advise them to bring it in so that I could see the effect. And so the lady and her husband turned up a couple days later with the set. Keen to see this strange phenomenon for myself, I turned the set on while they were there. And it promptly did all the right things; it brought up the channel the customer selected and it stayed there. No problems – except that, once again, I was stuck with a problem that I couldn’t see. So what if the channel had been selected via the remote control? They weren’t sure and they hadn’t brought the remote control with them. That meant the job had to be put on hold until they could drop it in. So I put the set in a corner and ran it for the rest of the day and for a couple hours the next morning, with no signs of trouble. When the customer came in with the remote control unit, it initially wouldn’t work at all. The reason was simple enough; the batteries were flat. So I fished out new batteries and fitted This view shows the innards of the remote control unit for the National TC-2178 TV receiver. At left is a rear view of the rubber pad & buttons, while at right is the PC board showing the button con­tact areas, the IC pins (bottom) & the IR LED at top. them, working at the bench with my back to the set. When I turned around, it had a white screen – it had moved to the blank position 11. At the customer’s suggestion, I pushed a button for one of the other channels and this appeared immediately. But then, as soon as I released it, we were back to position 11 – just as the customer had said. Well, at least I’d seen the fault and I sent the customers on their way while I thought things out. And it didn’t take much thinking to work out a likely theory; the set had behaved per­fectly until the remote control appeared on the scene. So the fault had to be in that, rather than in the set. It was easy enough to prove. I moved the remote control out of range and selected another channel at the set. It behaved perfectly. But as soon as I brought the remote control unit within range, we were back to position 11. So the control unit was transmitting a position 11 signal continuously – which also accounted for the flat batteries. Well, remote control faults are fairly common but this was a new one on me. Remotes are vulnerable Most remote control faults involve abuse of some kind. By their very nature, they are vulnerable devices. They are sat on, dropped, kicked, trodden on and generally bashed around. They are also soaked in various liquids – coffee, water, lemonade, beer and any other beverage you can think of. Liquid abuse, with the possible exception of water, means a write-off. And even water needs to be treated promptly, to ensure any chance of success. Otherwise, you drop it in the rubbish bin. But there ware no signs of abuse in this case. In fact, both the set and the remote control had been kept in immaculate condition. So my first guess was that button 11 was faulty, locking on in some way. February 1994  53 SERVICEMAN'S LOG – CTD By releasing one screw above the battery compartment and easing a knife blade between the two halves of the case, I was able to get the back off and lift out the PC board. Apart from the normal component connections, this carries copper contacts which sit behind the front panel buttons. The buttons have a conductive surface and are held in a rubber-like pad. When a button is pressed, it connects with its appropriate contact on the PC board. I could find nothing wrong here. The button was not jammed, and there was no foreign matter between the pad and its matching contacts. More importantly, the device was still generating a position 11 signal, even with the pad and PC board separated. It didn’t take much effort to narrow this fault to the IC, which is really the heart of the device. A new IC? No way; not available. That left a new remote control as the only option but it wasn’t a very satisfactory one; the price is around $150 when it is avail54  Silicon Chip able – which it wasn’t, stocks being on back order. In fact, the price structure on these devices is hard to understand. They are all basically the same – though seldom compatible – and yet prices range from around $50 to $150. Cold comfort And so it was all cold comfort for the customer. While I assumed they would accept the situation and pay up, I regretted that I could do nothing better. But then, while I was actually writing these notes, I sud­denly had a thought. Many months previously, I had been forced to write-off another National TV set as being uneconomical to re­pair. There had been a remote control unit with that, so what had happened to it? I moved immediately to my pre-loved, surplus equipment department and consulted the records (read: scrabbled through the junkbox). Sure enough, there it was. It had belonged to a more elabo­rate set, with many more remote control features (colour satura­tion, contrast, etc), but it also had the same basic functions as needed for this set. So would it work with it? I fitted a set of batteries and gave it a workout. Result – total compatibility with all the functions the set could provide. It would be a simple matter to ignore the other functions. In fact, this is not an unusual situation, even with new sets. In some cases sets are sold with a remote control having, say, Teletext controls, even though the set has no Teletext facility. It is obviously aimed at rationalising production and doesn’t seem to worry anyone. So I was able to offer the customer two options: a new control unit at around $150, or a secondhand one at a fraction of that figure. Of course, they jumped at the chance for the cheaper solution. So I scored a happy customer and made a small profit on a piece of surplus gear. It was smiles all round. Finally, a likely explanation for the failure. There had been a number of storms around his area recently and I have been involved in repairing some of the resultant damage, which has been quite extensive. TV sets, VCRs, microwave ovens, remote con­trolled roller doors, CD players, electric clocks and telephones have all suffered. When I mentioned this, the customer recalled that they had lost their roller door control, a clock and some other minor appliance during the storms but the TV set and VCR had not suf­fered. However, now that I had raised the point, he realised that the position 11 problem had appeared at about this time. It may have been pure coincidence, of course, but we do know that solid state devices are particularly vulnerable to lightning strikes – not so much direct strikes but to strikes in close proximity, which can produce magnetic fields to which these devices are sensitive. So there it is; no positive proof but SC a likely explanation. AMATEUR RADIO BY GARRY CRATT, VK2YBX Build a 6-metre handheld transceiver Amateurs looking to experience the 6-metre FM band might care to consider this project. Just buy an inexpensive walkie-talkie & fit it with crystals for the 6-metre amateur band. As many amateurs are no doubt aware, there exists in the electronics marketplace a device called a “headset communicator”. This item, usually sold in pairs, has featured in US mail order catalogs for at least 10 years and has been sold in Australia by at least half a dozen outlets. The original concept was a VOX operated headset, which controlled a low-power FM transceiver, operating on 49MHz. The antenna was formed by a piece of thin spring steel wire, which protruded from the headset, giving the operator a space age appearance. Of course, the frequency of 49MHz played havoc with Channel 0 television reception and several years ago the then Department of Communications produced an appropriate technical specifica­tion, removing the devices to 55MHz and reducing the operating power to a non-interfering level. The appropriate specification is now called ECR-60 and allows a transmitter output of 2mW EIRP, which equates to a field strength of 99dB µV/m measured at three metres. Given these limitations, such a device can not really be considered as a serious piece of communications equip­ ment and, as expected, has a quite low retail price – around $70. However, given the characteristics of the 6m FM band during the hot summer months, when temperature inversions can cause excellent signal propagation, such a device could form the basis of a simple FM handheld transceiver. Fortunately, Dick Smith Electronics carry a version of this style of FM transceiver which looks like any other handheld transceiver rather than the ridiculous looking headset version. R29 1k C34 .02 JUNCTION OF T1 AND R2 R27 15k C65 R26 6.8k +4.2V C36 0.1 T3 R50 4.7k BASE, Q2 Q3 9018G TO R51 R48 X3 68k C35 R28 680  +8V C57 0.2 Q12 9018G C56 C55 D10 Fig.1: this is the first local oscillator in the unit. Crystal X2 should be changed to 41.825MHz. 56  Silicon Chip R49 100  L1 C8 X2 Branded Digitor, the unit sells for just $69.95 (Cat. No. D 1095). Best of all, the unit is fitted with crystals for 55.150MHz, which is not too far from our target frequency of 52.525MHz, the national FM calling frequency. Our inspection of the workings of one of these units re­ vealed a dual conversion receiver using an MC3361 IF strip, an 8-pin DIL audio amplifier, a dual op amp for TX PTT and VOX func­tions, and some TX/RX switching diodes feeding what we assume is a loaded 55MHz helical antenna. The transmitter uses a 44MHz fundamental crystal, while the receiver uses an 18MHz third overtone crystal. The receiver first IF is 10.7MHz and the second IF is 455kHz, the expected conven­tional arrangement. We decided to put the unit on 52.525MHz, the national FM calling frequency, due to the difficulty in finding an accurate list of 6m FM repeaters. Whilst many repeaters are licensed to operate, few seem to exist. However, the unit would seem perfect­ly suitable for repeater operation. The mathematical calculation for 52.525MHz followed logi­cally: divide BASE, Q13 C54 L9 1.5uH Fig.2: this is the transmitter oscillator. Crystal X3 should be changed to 17.508MHz. X3 L1 T3 VR4 X2 L5 T1 T2 X1 VR1 Fig.3: this photo of the Digitor 55MHz transceiver shows the positions of the principal components, some of which may need to be adjusted for best performance after the crystals have been changed. 52.525 by three to obtain the Tx crystal frequency; ie, 17.508MHz. To obtain the Rx crystal frequency, subtract the first IF (10.7MHz) from the carrier frequency (52.525MHz) – in this case 41.825MHz. A call to Darren McCloud at HY-Q Crystals revealed that they had on file the exact specifications for both crystals, having been approached by various importers over the years for sample units. For the record, the transmitter specification is GG05S and the receiver specification is GG05Q. Both crystals are housed in the standard miniature wire in QC-49 holder. The next step was to replace the existing crystals with the HY-Q replacements and retune the transceiver. The photo of Fig.3 shows the PC board of the transceiver. Marked on the photo are crystal X2 (the receiver crystal) and X3 (the transmitter crystal). The remaining crystal, X1 has a frequency of 10.245MHz and is the second IF mix-down crystal. Also marked on the photo are VR1 (the preset mute trimpot), inductor T3 (the local oscillator coil) and other components which need to be adjusted during the align­ment procedure. Minor re-alignment is required to obtain optimum perfor­mance on the new operating frequency. Receiver alignment is best done with the receiver unmuted. Potentiometer VR1 adjusts the preset mute level and can be rotated anticlockwise (looking from the edge of the PC board) to unmute the receiver. A signal gen­erator should be used to re-align the receiver front end. The best connection point is where the existing helical antenna is connected to the PC board. It may be necessary to run a level of 10µV or so for initial alignment, backing the signal generator off as the receiver sensitivity improves with tuning. Inductor T3 adjusts the receiver local oscillator. The frequency can be measured at the base of transistor Q2, or the audio output can be monitored whilst adjusting T3. Receiver sensitivity can be improved by adjusting T1 and T2. As the transmitter only contains three devices, alignment is simple. Inductor L1 adjusts the output frequency and L5 adjusts antenna matching, best checked by monitoring field strength during adjustment. There is also a mysterious phase cancelling coil, having a few turns wrapped around the loaded helical anten­ n a. Dick Smith Electronics advises that the purpose of this is to reduce second and third harmonics when operating the unit with a headset. The headset wiring apparently changes the loading of the antenna, and subsequently the radiated harmonic output. A possible modification we considered was replacing the “mic sens” pot on the top panel of the unit with a fixed resistor, and using the redundant potentiometer to control the receiver mute. This would save running the receiver permanently either muted or unmuted, the only two options with the existing trimpot arrange­ment. The transmitter audio has two adjustments, VR2 adjusts the deviation, whilst front panel control VR3, a 10kΩ potentiometer, adjusts the microphone sensitivity. This adjustment is really designed to have the effect of changing the VOX sensitivity. We also considered replacing the existing helical antenna with a coaxial “tail” terminated with a BNC socket. This would allow connection of an external antenna. No doubt there are other modifications which can be made to the unit to further improve performance. In any case, relocating what might otherwise have been dismissed as a “toy” transceiver to the 6m band will prove to be a worthwhile exercise for many amateurs. Finally, HY-Q Crystals has advised that they can supply crystals at around $30 each. They can be contacted by SC phone on (03) 562 8222. February 1994  57 Ma • High in Featu • 50-h intensity am res duty c our battery ber LED life (A A hea • 1 0 0e- lls) vy h o ur ba alkalin tter y e cells l i fe ( • Com ) AA • Conspact size t a nt LED batter bright ness o • Battye life ver ry con dition indica tor Novel LED torch has low battery drain Using a highly efficient amber LED, this LED torch has the advantage of small size & prolonged battery life. In addi­tion, you will never need to replace the “lamp”. By JOHN CLARKE This torch is not as bright as, say, a conventional penlite torch but it has much better battery life. It’s great for finding keys in a handbag, lighting up a keyhole when opening your front door at night or any time you don’t want an ordinary torch which is really too bright for the job. There are many instances where ordinary torches are just too bright. After all, we do not always want to spot possums in the trees or ward off intruders. In fact, the lower light output from a LED torch is useful when 58  Silicon Chip checking on a sleeping child at night and for use by astronomers who don’t want to disturb their “dark-adapted” eyes. A big problem with ordinary torch­ es is that their batteries always seem to be on their last legs when you want to use them. With this LED torch, you can expect up to 15 times the battery life of a standard penlite (two AA cells) torch. These torches typically draw 300mA from the battery while this LED torch only draws about 25mA. The gain in battery life is partly due to the lower current drain and partly to better battery efficiency at lower currents. The idea of a using LEDs in a torch has been around for some time but until recently, suitable LEDs were not available. Grant­ed, high intensity red LEDs can be used but the red colouring is not pleasant. Since there is no such thing as a white LED, the new high-brightness amber LEDs are the go. Specifically, the new Hewlett Packard AlInGaP (Aluminium Indium Gallium Phosphide) LEDs are preferred for this job. The amber light corresponds to the more sensitive spectrum region of our eyes and it gives better render­ing of the colour of objects. We built the LED torch into a small plastic case. The LED is mounted on one end of the case while a slide switch on the top turns it on and off. In the base is a battery condition indicator. A OFF CELL1 220 16VW LED1  K S1 ON LAMP1 CELL2 A K LED TORCH Fig.1: the circuit consists of two AA cells which drive a high-brightness LED via a series 1.5V lamp. The lamp ensures constant LED current. It glows brightly when the battery is good but gradually dims over the life of the battery, eventually ceasing to glow when the battery is at the end of its life. Circuit details There is not much to the circuit although there is more than you might expect. In the simplest arrangement, we could have had two AA cells feeding the LED via a resistor selected to set the current at around 25 milliamps. This works but has the disadvan­tage that the LED will gradually dim as the batteries age. Hence, we have used a slightly more exotic circuit involving a series 1.5V incandescent lamp and a 220µF 16VW capacitor. The lamp ensures that the current through the LED will remain relatively constant over the life of the battery. It operates on the principle that the resistance of a light bulb increases with the filament temperature. So when the batteries are new, the light bulb will have almost 1.5V across its filament but when the batteries are old, there is almost no voltage lost across the filament. Just how well the current regulation works can be judged by comparing it with a LED driven directly via a resistor. As the batteries age, their total The PC board is dominated by the two AA cell holders. Make sure that the LED & the 220µF electrolytic capacitor are correctly oriented. voltage will range from 3.3V down to about 2.2V and this will result in a reduction of LED current of more than 70%. The LED/lamp system, by comparison, results in a current reduction of just over 40% for the same voltage range. But that is not the end of the story. With the LED/lamp system, the LED will continue to put out useful light when the battery voltage has diminished to 1.9V; ie, 0.95V per cell. The 220µF capacitor is included to prevent surge current through the LED when power is first applied. This would otherwise occur due to the low cold resistance of the lamp bulb. Note that the capacitor is shorted each time the slide switch is turned off to ensure that it is discharged before power is reapplied. Construction To make assembly easy, the parts are mounted on a PC board measuring 79 x 41mm and coded 08302941. You can begin construction by clipping off the corners of the PC board so that it will fit neatly into the case without fouling PARTS LIST 1 PC board, code 08302941, 79 x 41mm 1 plastic case, 24 x 50 x 90mm (Jaycar HB-6031) 2 adhesive labels, 25 x 8mm 1 SPDT slider switch (C&K 1101 or equivalent) 2 AA cell holders with flying leads or solder terminals (Altron­ics S-5026 or Tandy 270-401) 1 1.5V 25mA mini lamp with grommet (Tandy 272-1139) 1 5mm Hewlett Packard amber LED, HLMA-DL00 (VSI) 1 5mm LED bezel 1 220µF 16VW PC electrolytic capacitor 4 2.5 dia. x 5mm machine screws & nuts 4 3 dia. x 5mm machine screws 3 PC stakes 1 50mm length 0.8mm dia. tinned copper wire 220uF CELL 1 A LED1 LAMP1 K CELL 2 S1 Fig.2: install the parts on the PC board as shown here. Fig.3: the full-size etching pattern for the PC board. February 1994  59 The 1.5V lamp protrudes through a hole in the PC board & is shock-proofed by fitting it with a rubber grommet. the corner pillars. You may also need to drill out the hole for the light bulb grommet. Install the PC stakes, capacitor and single cell battery holders, making sure that the latter parts are correctly oriented. The battery holders are secured to the PC board with 2.5mm screws and nuts. Note that you cannot use AA battery holders which have clips at each end since they will be too long for the case. Install the grommet and wire the bulb in place. This done, temporarily mount the switch on the PC stakes by soldering to the centre pin only. This will allow easy adjustment later. Drill a hole in the centre of one of the case ends to fit the LED bezel and solder the LED in place on the board, as shown in the photograph. You can also drill a small hole in the base of the case for the lamp to shine through, for battery indication. Secure the PC board to the case using four 3mm screws. Now you will OFF ON LED TORCH Fig.4: here are the full size artworks for the two adhesive labels. need to drill and file out a rectangular hole in the lid of the case for the switch slider. Adjust the switch height and position so that it can be operated freely when the case is assembled. Finally, solder the remaining PC stakes to the switch leads. We made up a couple of small labels for the lid and base of the case. If you have these, roughen the surface of the case with emery paper to allow them to stick properly. Install the batteries and check that the torch operates. If not, check the polarity of the LED. Finally, assemble the case with the self-tapping screws. Note that the end pieces of the case fit properly only one way around even though they appear to be able to go in SC either way. High-Brightness LED Options While we recommend the HP HLMA-DL00 LED which has a 30° beam and 3001000mcd output at 20mA, there are two other Hewlett Packard amber LEDs which may be more suitable for your application. (1). HLMA-CL00 is also 5mm in diameter and similar to the DL00 except that it has a narrower beam of 7° and 1000-3000mcd at 20mA. It is more useful as an inspection light. (2). HLMA-BL00 is 13.3mm in diameter with a 4° viewing angle and with a higher intensity of 15 candelas at 20mA. This costs about $20 and has a powerful but very narrow beam. These LEDs are available from VSI Electronics Australia Pty Ltd in your capital city. Alternatively, the HLMA-DL00 and HLMA-CL00 are available from Farnell Electronic Components. Phone (02) 645 8888. 60  Silicon Chip SILICON CHIP If you are seeing a blank page here, it is more than likely that it contained advertising which is now out of date and the advertiser has requested that the page be removed to prevent misunderstandings. Please feel free to visit the advertiser’s website: www.jaycar.com.au SILICON CHIP If you are seeing a blank page here, it is more than likely that it contained advertising which is now out of date and the advertiser has requested that the page be removed to prevent misunderstandings. Please feel free to visit the advertiser’s website: www.jaycar.com.au SILICON CHIP If you are seeing a blank page here, it is more than likely that it contained advertising which is now out of date and the advertiser has requested that the page be removed to prevent misunderstandings. Please feel free to visit the advertiser’s website: www.jaycar.com.au SILICON CHIP If you are seeing a blank page here, it is more than likely that it contained advertising which is now out of date and the advertiser has requested that the page be removed to prevent misunderstandings. Please feel free to visit the advertiser’s website: www.jaycar.com.au ORDER FORM BACK ISSUES MONTH YEAR MONTH YEAR PR ICE EACH (includes p&p) Australi a $A7.00; NZ $A8.00 (airmail ); Elsewhere $A10 (airmail ). Buy 10 or more and get a 10% discount. Note: Nov 87-Aug 88; Oct 88-Mar 89; June 89; Aug 89; Dec 89; May 90; Aug 91; Feb 92; July 92; Sept 92; NovDec 92; & March 98 are sol d out. All other issues are currently i n stock. TOTAL $A B INDERS Pl ease send me _______ SILICON CHIP bi nder(s) at $A12.95 + $5.00 p&p each (Australi a only). N ot avail abl e elsewhere. Buy five and get them postage free. $A SUBSCRIPTIONS ❏ New subscription – month to start­­___________________________ ❏ Renewal – Sub. No._______________   ❏ Gift subscription ☞ RATES (please tick one) Australia Australia with binder(s)* NZ & PNG (airmail) Overseas surface mail 2 years (24 issues) 1 year (12 issues) ❏ $A90 ❏ $A49 ❏ $A114 ❏ $A61 ❏ $A135 ❏ $A72 ❏ $A135 ❏ $A72 ❏ $A240 Overseas airmail ❏ $A120 *1 binder with 1-year subscription; 2 binders with 2-year subscription GIFT SUBSCRIPTION DETAILS Month to start__________________ Message_____________________ _____________________________ _____________________________ Gift for: Name_________________________ (PLEASE PRINT) YOUR DETAILS Your Name_________________________________________________ (PLEASE PRINT) Address___________________________________________________ Address______________________ _____________________________ State__________Postcode_______ ______________________________________Postcode___________ Daytime Phone No.____________________Total Price $A __________ ❏ Cheque/Money Order ❏ Bankcard ❏ Visa Card ❏ Master Card 9am-5pm Mon-Fri. Please have your credit card details ready ______________________________ Card expiry date________/________ Card No. Phone (02) 9979 5644 Signature OR Fax (02) 9979 6503 Fax the coupon with your credit card details 24 hours 7 days a week Mail coupon to: OR Reply Paid 25 Silicon Chip Publications PO Box 139, Collaroy 2097 No postage stamp required in Australia February 1994  65 Build a 40V 3A variable power supply This month, we complete the 3A-40V Adjustable Power Supply by describing the construction, testing & setting up pro­cedures. Most of the parts mount on a large PC board, so the assembly is straightforward. PART 2: By JOHN CLARKE 66  Silicon Chip A large PC board coded 04202941 (222 x 160mm) carries the bulk of the electronic circuitry, including the power transform­ er. This board is mounted on pillars moulded into the base of the case and secured using self-tapping screws. Most of the remaining parts are mounted on the front panel and are connected to the PC board via insulated leads. Board assembly Fig.9 shows the parts layout on the PC board. Begin by checking the board Fig.9 (facing page): install the parts on the PC board as shown on this combined layout & wiring diagram. The leads marked with an asterisk (*) must be run using 32 x 0.2mm wire in order to carry the heavy currents involved. ▲ The S ILICON C HIP 3A-40V Adjustable Power Supply is housed in a standard plastic instrument case measuring 260 x 190 x 80mm. This is fitted with aluminium front and rear panels, the rear panel providing the necessary heatsinking for the switching regulator (IC1). In addition, these aluminium panels are connected to the mains earth to ensure safety and play an important role in shielding the supply circuitry. Do not, under any circumstances, use plastic panels for this project. for etching defects by comparing it with the published pattern. Usually there will be no problems but it’s always best to make sure before mounting any of the parts. If everything is OK, start the assembly by installing PC pins at all external wiring points, then install the resistors and wire links. Table 1 lists the resistor colour codes but it’s best to also check them on your multimeter as some of the colours can be difficult to decipher. Note that the two 680Ω 5W resis­tors should be mounted about 1mm above the board to allow air circulation, while the 4.7kΩ resistor ACTIVE (BROWN) FUSE EARTH (GREEN/YELLOW) METAL REAR PANEL EARTH TERMINALS CORD GRIP GROMMET GREEN/YELLOW GREEN/YELLOW 1 IC1 D1-D4 100uF 1000uF 4700uF 680  5W D5 NEUTRAL (BLUE) 4700uF 22 1000uF 1.5k 21 L1 PRI 15k 100  VR3 2.2k 680  330pF 15V 0V VR4 1k 10k 100k 47k 91k  15 16 17 1k 2.2k 10k IC5 4053 14 10k 1k 1 D6 1k 47k 220  22k 0.1 0.1 100  IC3 LM339 POWER TRANSFORMER 1k 1M 6.8k 1 4.7k IC2 OP77 15V 0V 100uF IC4 OP77 ZD2 REF1 1 10uF 0.1 10uF 100uF ZD1 1  IC6 7660  L2 R1 0.1 680  5W 18 19 20 100k 10uF 1 13 12 11 10 9 8 7 6 5 4 3 2 0.1 250VAC 0.33 GND  SEE TEXT GREEN/YELLOW S1 10 13 12  17    S4 S2  9 8 7 6 1 5 4 15 16 3  14 GND 11 22 2 I/P 7106 DPM-02 VR2 20 A A K LED1 METAL FRONT PANEL 19 18 BATT S3 21 VR1 K LED2 SOLDER LUG ON POT BUSH February 1994  67 The switching regulator (IC1) is bolted to the rear panel for heatsinking but must be isolated from the panel using an insulating bush & washer. A separate TO-220 style heatsink is fitted to diode D5. The connections to the LCD panel meter are made by soldering leads to the terminals on the back of the PC board. Use a small fine-tipped soldering iron for this job. A few dabs of epoxy resin can be used to hold the panel meter in place. adjacent to zener diode ZD2 must be rated at 0.5W. The link designated R1 must be run using 0.4mm diameter enamelled copper wire (note: this is the current sense resistor). Tin each end of the link (scrape away the enamel at each end first) before mounting it on the PC board. This will ensure a good solder joint at each end of the link. Do not use any other type of wire for this link, otherwise you will have trouble calibrating the supply later on. 68  Silicon Chip Next, install the ICs, zener diodes, diodes, REF1 and the trimpots. Solder only the two outside pins of IC1 at this stage (do not trim the leads) so that it can be later easily adjusted to line up with its mounting hole in the rear panel. Make sure that the ICs and diodes are correctly oriented and be sure to use the correct part number at each location on the board. Zener diode ZD1 should be mounted with a small loop in one end to provide thermal stress relief. Diode D5 is mounted on a small TO-220 style heatsink fitted with two locating lugs. Smear the metal tab of the diode with heatsink compound, then bolt it directly to the heatsink using a machine screw and nut (no mica washer necessary). The resulting assembly can then the fitted to the board and the leads soldered. Note that the locating lugs on the heatsink go through two matching holes in the PC board. Bend these lugs slightly to secure the heatsink in place. The capacitors can now all be installed on the PC board but watch the polarity of the electrolytic types. Take care when installing the three 100µF electrolytic capacitors; two of these are rated at 63VW while the third is rated at just 16VW. The latter is installed adjacent to ZD1. Winding the transformers Inductors L1 and L2 can now be wound and installed on the PC board. L1 is made by winding 50 turns of 0.8mm enamelled copper wire on its plastic bobbin former. Begin by pre-tinning one end of the wire and soldering this to terminal 10. This done, wind on the first layer (with each turn adjacent to the other) and cover it with a single layer of insulation tape. The remaining layers are then wound in exactly the same manner until 50 turns have been made, with each layer covered by a single layer of insulating tape. When the 50 turns are on, solder the wire end to terminal 4 and wind a couple of layers of tape over the completed windings. Before assembling the transformer, the centre leg on one of the ferrite core halves must be filed down so that there is a 1mm gap between the centre cores. You will need a flat file for this job – keep the file square to the ferrite core surface to main­tain an even gap across the entire face. A short length of 1.0mm-diameter wire is used as a feeler gauge to check the gap at regular intervals. When the gap is correct, the cores can be inserted into the bobbin and the metal retaining clips snapped in place. L2 is wound on a toroid former using two 1-metre lengths of 1.5mm enamelled copper wire – see Fig.10. There are two separate 14-turn wind­ ings, L2a and L2b, and these must be wound in the directions shown to ensure correct phasing. Wind the turns on firmly and strip and tin the Fig.10: inductor L2 is made by winding two separate 14-turn coils on a toroid former. Wind the coils exactly as shown here, to ensure correct phasing. wire ends to ensure good solder joints to the PC board. L1 and L2 can now both be installed as shown in Fig.9. Note that a plastic cable tie is used secure L2. Finally, transformer T1 can be secured to the board using 4mm screws, washers and nuts. Preparing the case Some of the integral pillars on the base of the case must be removed in order to accommodate the PC board. Fig.11: the mounting details for IC1. Smear all mating surfaces with thermal grease before bolting the assembly together. To do this, first fit the board to the base and use a felt-tipped pen to mark its five mounting pillars (ie, the five directly beneath the board mounting holes). This done, remove the PC board and remove all the unused pillars using an oversize drill. The five remaining mounting pillars should also be cut down by about 1mm, so that the transformer will fit within the case when the lid is on. In addition, the case lid has a small raised bar running across its centre and this should be removed using side cutters or a sharp chisel. If you are building the power supply from a kit, the front and rear panels will be supplied pre-punched, while the front panel will also come with screen printed labelling. Altern­ative­ly, if you are starting from scratch, drill a mounting hole for two earth lugs in the top lefthand corner of the panel, then mount the two earth lugs using a countersunk screw plus nuts and washers (note: use a couter­sunk dress February 1994  69 Use plastic cable ties to lace the wiring together & make sure that none of the mains leads can come adrift & short against the case or other parts. The fuse & power switch (S1) are both covered with heatshrink tubing, to prevent accidental contact with the 240V AC mains. screw if the front panel is supplied screen printed). The front panel label can now be fitted and used as a drilling template for the various holes. It’s always best to drill small pilot holes first and then carefully enlarge them to size using a tapered reamer. The square cutouts for the LCD panel meter and for switches S1 and S2 can be made by first drilling a series of small holes around the inside perimeter of the marked areas, then knocking out the centre pieces and filing each cutout to shape. The DVM-02 module is initially held in the front panel by making it a force fit, so be careful not to make its cutout too big. A small dab of epoxy resin along each side of the module (applied from the back of the front panel) is then used to secure the LCD module in position. On the rear panel, you will need to drill holes to accept the mains fuse 70  Silicon Chip (F1), the cord grip grommet and three solder lugs. The wiring diagram (Fig.9) shows the locations of these holes. In addition, you will also have to drill a mounting hole for IC1. The location of this mounting hole can be determined by fitting the PC board inside the case and sliding the rear panel into position. Mark out and drill the hole, then carefully deburr it using an oversize drill so that the surface is perfectly smooth. Finally, refit the rear panel and adjust IC1 as necessary before soldering its three remaining pins to the PC board. Fig.11 shows how IC1 is isolated from the rear panel using a mica washer and insulating bush. Smear all surfaces with heat­sink compound before bolting the assembly together (note: heat­sink compound is unnecessary if you use one of the new silicone impregnated washers). Finally, check that the metal tab of IC1 is indeed isolated from the rear panel using a multimeter switched to a low ohms range. The PC board assembly can now be attached to the base of the case using five self-tapping screws and the various hardware items mounted on the front and rear panels – see Fig.9. Before mounting the potentiometers, cut the shafts to a length to suit the knobs and note that a large solder lug is fitted to the shaft of VR1. A similar large solder lug is also fitted to the GND output terminal. Important: if the aluminium panels are anodised, you will need to scrape away the anodising from around the earth lug holes to ensure good electrical contact. Final wiring Fig.9 shows the final wiring details. Begin this work by stripping back the outer insulation of the mains cord by 170mm, so that the leads can reach the mains switch (S1) on the front panel. This done, push the mains cord through its entry hole and clamp it securely to the rear panel using the cordgrip grommet. The Neutral (blue) mains lead goes directly to switch S1, while the Active (brown) lead goes to S1 via the fuse. Slide some heatshrink tubing over the leads before soldering the connec­tions. After the connections have been made, the tubing is shrunk over the switch and fuse to prevent accidental contact with the mains. The green/yellow striped lead from the mains cord connects directly to the rear panel earth using a crimp lug terminal. Additional green/yellow earth wires are then run from the rear panel earth to the front panel, from the front panel to the power transformer frame, and finally from the solder lug on VR1 to an earth terminal at top right on the rear panel. Note that the two earth leads running between the front and rear panels are critical in obtaining low residual hash in the supply output. Do not leave these leads out. Light-duty rainbow cable is used for wiring the LEDs, while most of the remaining leads are run using light-duty hook-up wire. The exceptions are those leads marked with an asterisk (*). These must be run using 32 x 0.2mm wire in order to carry the heavy currents involved (ie, to the transformer secondary termi­ nals, to the output terminals and to switch S2). Note that the heavy-duty leads running from near L2 on the PC board to switch S2 are twisted to prevent noise pick-up from the switchmode circuitry. Use plastic cable ties to The centre leg on one of the ferrite core halves used for L1 must be filed down so that there is a 1mm gap between the centre cores when the inductor is assembled. The photo below shows how the ferrite core is pushed into the plastic bobbin. lace the wires together, to give a neat appearance. In addition, use several plastic cable ties to lace the mains wires together. This is an important safety measure as it prevents any wire that may come adrift from making accidental contact with any part of the metalwork or vulnerable low-voltage circuitry. Be warned that the wiring to switch S4 may present a few problems if the switch specified in the parts list is not used. This is because some momentary pushbutton switches have their common (C) terminals in the middle and their normally open (NO) and normally closed (NC) contacts on the RESISTOR COLOUR CODES ❏ ❏ ❏ ❏ ❏ ❏ ❏ ❏ ❏ ❏ ❏ ❏ ❏ ❏ ❏ ❏ No. 1 2 1 2 1 1 3 1 1 2 1 5 1 1 2 Value 1MΩ 100kΩ 91kΩ 47kΩ 22kΩ 15kΩ 10kΩ 6.8kΩ 4.7kΩ 2.2kΩ 1.5kΩ 1kΩ 680Ω 220Ω 100Ω 4-Band Code (1%) brown black green brown brown black yellow brown white brown orange brown yellow violet orange brown red red orange brown brown green orange brown brown black orange brown blue grey red brown yellow violet red brown red red red brown brown green red brown brown black red brown blue grey brown brown red red brown brown brown black brown brown 5-Band Code (1%) brown black black yellow brown brown black black orange brown white brown black red brown yellow violet black red brown red red black red brown brown green black red brown brown black black red brown blue grey black brown brown yellow violet black brown brown red red black brown brown brown green black brown brown brown black black brown brown blue grey black black brown red red black black brown brown black black black brown February 1994  71 Fig.12: check your etched PC board against this full-size pattern before installing any of the parts. The board is coded 04202941 & measures 222 x 160mm. 72  Silicon Chip . (+) . (-) . GND . SET DROPOUT . . OVERLOAD POWER CURRENT LIMIT METER .V . A. . . 3A-40V ADJUSTABLE POWER SUPPLY Before applying power, carefully check your work for any wiring errors. This done, wind VR1 fully anticlockwise and set VR2, VR3 and VR4 to centre position. Switch on the supply and immediately check that the voltage across ZD1 is about 9V. If so, check the reading on the digital display. It should show about 1.23 volts if S3 (the Meter switch) is in the “V” position, or about 0.00 amps if it is in the A position (note: the least significant digit will be incorrect until VR4 is adjusted later on). If everything is OK at this stage, you can check the supply voltages to each IC. Connect your multimeter negative lead to the cathode of ZD1 and check the voltage at pin 7 of IC2 and IC4, pin 3 of IC3, pin 16 of IC5 and pin 8 of IC6. These should all be at +9V. Pin 4 of IC2 should be at about -9V. If at any stage the voltages are incorrect, switch off immediately and correct the problem before proceeding. The output voltage from the power supply should be adjust­ able from 1.23V up to about 43V, with the dropout LED lighting at about 42V (depending on mains voltage). Check that the voltage reading on the panel meter changes from 2-digit resolution after the decimal point to 1-digit resolution at 15-18V. When the panel meter is set to read amps, the display may initially read several digits above or below 0.00. This can be corrected by adjusting VR4. This done, set the Current Limit control (VR2) fully anticlockwise and press the Set switch (S4). Check that the display still reads 0.00 – if not, adjust VR4 accordingly (the adjustment will only be slight). Now press the current set switch and check that the display reading can be varied from 0.00 up to at least 4.00A by adjusting the Current Limit control. Note that the overload LED may light when the control is fully anticlockwise. This is normal and the LED will extinguish when the current limit reaches 10mA (0.01 on the display). When measuring voltage, the panel meter should be accurate to 1% without calibration. However, if you have an accurate voltmeter, the trimpot on the back of the DVM-02 can be adjusted to give even greater accuracy if required. For current readings, the panel meter is calibrated by first connecting a 4.7Ω 5W resistor across the output and setting the supply to deliver 4.70V. The Current Limit control should now be rotated at least half-way, to prevent the current limit fea­ture from operating. This done, switch S3 to the “A” position and adjust VR3 until the meter shows 1.00 amps. Warning – the resistor will become quite hot during this procedure. The current limiting feature should now be checked for correct operation. To do this, leave the 4.7Ω resistor in circuit and rotate the Current Limit control anticlockwise until the overload LED lights. This should initially occur at 1A but you should now be able to set lower current limits by further reducing the control setting. The power supply will squeal during current limiting but this is normal. Finally, you can check the power supply on various loads and if you have access to an oscilloscope, you can observe the SC output ripple. LOAD Testing Fig.13: this full-size artwork can be used as a drilling template for the front panel. If you buy a kit, the panel will be supplied pre-punched & screen printed. . VOLTAGE ADJUST . outside, whereas the switch we used has its common terminals at one end. If your switch has its common terminals in the middle, the wiring shown in Fig.9 will no longer be relevant and you will have to work out the connections from the circuit diagram (Fig.5). The common, NO and NC terminals will usually be marked somewhere on the body of the switch. February 1994  73 SILICON CHIP If you are seeing a blank page here, it is more than likely that it contained advertising which is now out of date and the advertiser has requested that the page be removed to prevent misunderstandings. Rod Irving Electronics Pty Ltd SILICON CHIP If you are seeing a blank page here, it is more than likely that it contained advertising which is now out of date and the advertiser has requested that the page be removed to prevent misunderstandings. Rod Irving Electronics Pty Ltd SILICON CHIP If you are seeing a blank page here, it is more than likely that it contained advertising which is now out of date and the advertiser has requested that the page be removed to prevent misunderstandings. Rod Irving Electronics Pty Ltd SILICON CHIP If you are seeing a blank page here, it is more than likely that it contained advertising which is now out of date and the advertiser has requested that the page be removed to prevent misunderstandings. Rod Irving Electronics Pty Ltd SILICON CHIP If you are seeing a blank page here, it is more than likely that it contained advertising which is now out of date and the advertiser has requested that the page be removed to prevent misunderstandings. Rod Irving Electronics Pty Ltd COMPUTER BITS BY DARREN YATES Experiments with your games card; Pt.4 This month, we look at the games card port & learn how each bit is defined. We also discuss how you can use the BIOS interrupt routines to get fast information about the port. In the January 1994 issue, we looked at how you can tell whether or not a PC has a games card installed (apart from having a look at the back). This was done using a machine code routine inside a QBASIC program. A similar process is used by games programs to find out this information. So far we’ve spent some time on the analog inputs which read the X and Y coordinates of a joystick. (Remember that the card can handle two joysticks at once). However, we haven’t covered the four digital inputs which are accessible via the fire buttons of the joysticks. These digital inputs are quite easy to use – in fact, much easier to use than the printer port inputs. Let’s take a look at the pinout diagram for the joystick DB15 socket – see Fig.1. As you can see, the four fire buttons in the joysticks (S1-S4) simply pull their corresponding inputs to ground. We don’t need to worry about maintaining 5V logic lines or anything else – we can simply pull each line to ground or leave it open and we can do this with a single transistor. Bits 7 to 4 are initially set to ‘1’ and become ‘0’ when the joystick button is pressed. You can test this by soldering two wires to pins 10 and 12 of the joystick adaptor plug and joining them together while running this short program in QBASIC. WHILE A$=”” A=INP(&H201) PRINT A A$=INPUT$(1) WEND You should see the number ‘243’ flash down the lefthand side of your screen and whenever you join the wires together, the number should change to ‘211’. The numbers themselves are not important but you should see the number change every Games card port Just as your printer card and serial card both have their own port addresses (the printer port is usually 0378H and the serial port 03F8H), so too does your games card and its address is 0201H. If we have a look at Table 1, we can see how each of the 8 bits is used. time you join the two wires. Alternatively, you could plug in a joystick if you have one handy, and run the same program while pressing the fire buttons. You should find that a similar thing happens – the ‘243’ number on the screen should change. What it changes to will depend upon which input your joystick is connect­ed to – either A or B. The four least significant bits each determine (in a way) the coordinate from the joysticks. This probably won’t be obvious from the outset but they work like this. Remember how we looked previously at the 558 timer circuit and how the joystick control formed part of the monostable cir­cuit? Well, the output of each monostable appears at one of these bits. Now the whole idea is that while the bit remains high, a counter should be counting elsewhere in the program to keep track of the time. When the bit goes low, the count represents a pro­portional figure to the joystick position. As you’ll probably agree, while using this port for the digital inputs is fairly straightforward, there is quite a bit more work to be done on the data from this port in order to obtain the analog inputs. If you think about your favourite flight simulator, just imagine how much calculation has to go on for the joy­stick alone to figure out where you are! BIOS interrupt Fig.1: the pin connection details for the DB15 sockets on a games card. Thankfully, this is where the BIOS interrupt routines come in - in particular INT 15H, SERVICE 84H. Table 2 shows how this works. Last time, we were able to obtain hardware information about the games card by using INT 11H and we can February 1994  79 Table 1: Game Adapter AB Joystick Data Byte Protect your valuable issues Silicon Chip Binders Bit Number 7 6 5 4 3 2 1 0 ✔ Status of B joystick button 2 ✔ Status of B joystick button 1 ✔ Status of A joystick button 2 ✔ Status of A joystick button 1 ✔ B joystick Y coordinate* ✔ B joystick X coordinate* ✔ A joystick Y coordinate* ✔ These beautifully-made binders will protect your copies of SILICON CHIP. They feature heavy-board covers & are made from a dis­ tinctive 2-tone green vinyl. They hold up to 14 issues & will look great on your bookshelf. ★ High quality ★ Hold up to 14 issues ★ 80mm internal width ★ SILICON CHIP logo printed in gold-coloured lettering on spine & cover Price: $A14.95 (includes postage in Australia). NZ & PNG orders please add $A5 each for postage. Not available elsewhere. Silicon Chip Publications PO Box 139 Collaroy Beach 2097 Or fax (02) 979 6503; or ring (02) 979 5644 & quote your credit card number. ➦ Use this handy form Enclosed is my cheque/money order for $________ or please debit my ❏ Bankcard   ❏ Visa   ❏ Mastercard Card No: ______________________________ Card Expiry Date ____/____ Signature ________________________ Name ___________________________ Address__________________________ __________________ P/code_______ 80  Silicon Chip Function A joystick X coordinate* also obtain the joystick coordinates as well as the button inputs using this interrupt routine. Interrupt routines Before we dive headlong into more machine code, the idea of interrupt routines may well be new to some readers, so we’ll take a more leisurely approach. Looking at Table 2, this routine, like most of the BIOS and DOS interrupts, uses the four general purpose 16-bit registers inside the 8086/8088/80286/386/486 processor; ie, AX, BX, CX and DX. Each of these can be split in half to give two 8-bit reg­isters – high and low. For the accumulator register AX, we can access the high eight bits by referring to AH and the low eight bits by referring to AL. The other three 16bit registers give, respectively, BH, BL, CH, CL, DH and DL. Remember that there are only four registers but they can be split or, more accurately, ‘selected’ as eight bits wide by using the ‘H’ and ‘L’ suffixes. In the 80386/486 processors, the AX to DX registers are themselves the lower 16-bit ‘splits’ of 32-bit registers EAX through to EDX, the ‘E’ prefix standing for ‘extended’. It’s not possible though to access the higher 16-bit sections of each of these extended registers. If we look back at Table 2, before the interrupt can be called using the instruction INT 15H, we have to load the number 84H into the upper 8-bit section of AX, namely AH. This tells the PC that we want a particular service out of those available from INT 15H. You can think of this service number as a house number and the interrupt number as a street name. We still have a further choice to make and that deals with the amount of information returned. By setting the DX register to ‘0’, the only information returned from the interrupt routine is just the joystick fire button settings and these are returned in Table 2: Joystick Support (Interrupt 15h, Service 84h) Registers on entry: AH:84h DX: 00h = read switches 01h = read joystick position Registers on Return: If reading switches (DX=0): AL = switch settings (bits 4-7) If reading position (DX=1); AX=A(X) value BX=A(Y) value CX=B(X) value DX=b(Y) value Memory affected: None Note: This service is not available on PC machines released prior to 1983. register AL. However, if we set DX to ‘1’, then register AX returns the X-axis coordinate from joystick A, BX the y-axis, CX the x-axis from joystick B and DX the y-axis. Note that the joystick button settings are ignored in this case. One good thing about this routine is that it returns all four coordinates at once, giving you the possibility of having four analog inputs sampled at the same time. Example OK, let’s take our new-found know­ ledge and write a short program. Let’s keep it simple and just return the settings of the fire buttons. Remember that these appear in register AL. The program, called BUTTON.BAS, uses a machine-code program inside QBASIC to get the information we want. If you’ve been following this series, you’ll notice the similarity between this program and the Games Card Finder program presented in the January 1994 issue. Going through it briefly, the machine code program is stored in the integer array ASMPROG. Lines 3 and 4 of the machine code load 84H into register AH and 0H into DX. After that, the INT 15H call is made. Again, the VARPTR and VARSEG provide the specific address information of the first element in the program array. This is so control is transferred to the correct position and so the program starts and runs correctly. The joystick fire button information is returned in the variable BUTTON. Each fire button bit is then separated out and passed to the BUTTON array from 1 to 4. As the comments in the program show, bit 7 from port 201H corresponds to button 1 which gives the value 128 if the button is not pressed, bit 6 corre­sponds to button 2 which gives 64 and so on – down to 32 and finally a value of 16 in BUTTON(4). If any of these buttons have been pressed however, the corresponding BUTTON array will give a value of ‘0’. This information is then printed on the screen for each of the four buttons. The FOR..NEXT loop cuts down on the program lines and simply checks each BUTTON array in turn, calculating what the correct number should be in that array element if that button wasn’t pressed. Button.Bas: Joystick Button Finder Program ‘Joystick button finder ‘Copyright 1993 SILICON CHIP ‘Written by DARREN YATES B.Sc. ‘ This program uses the BIOS interrupt 15, service 84 to obtain ‘ the status of the four fire buttons without using the BASIC ‘ commands. DEFINT A-Z DIM ASMPROG(1 TO 10) DIM button(1 TO 4) ‘The machine-code program is stored in the array ASMPROG and read ‘and read into the array. ASMBYTES: DATA &h55 : ‘PUSH BP save base pointer DATA &h8b,&hec : ‘MOV BP,SP get our own DATA &hb4,&h84 : ‘MOV AH,84H set service number DATA &hba,&h00,&h00 : ‘MOV DX,0000 select button input data only DATA &hcd,&h15 : ‘INT 15H make ROM-BIOS call DATA &h8b,&h5e,&h06 : ‘MOV BX,[BP+6] get argument address DATA &h88,&h07 : ‘MOV [BX],AL save list in argument DATA &h5d : ‘POP BP pop argument off stack DATA &hca,&h02,&h00 : ‘RET 2 and make far return to BASIC ‘get the starting offset of the array start = VARPTR(ASMPROG(1)) ‘poke machine code program into the array ASMPROG DEF SEG = VARSEG(ASMPROG(1)) RESTORE ASMBYTES FOR index = 0 TO 18 READ byte POKE (start + index), byte NEXT index ‘run the machine-code program start = VARPTR(ASMPROG(1)) CALL absolute(button, start) DEF SEG ‘variable BUTTON now contains info on the joystick buttons button(1) = button AND &H80 button(2) = button AND &H40 button(3) = button AND &H20 button(4) = button AND &H10 PRINT STRING$(18, 205) ‘this section selects the correct bit for each button ‘ bit 7 = button 1; bit 6 = button 2; bit 5 = button 3; bit 4 = button 4 ‘ if that bit is 0 then button is pressed FOR number = 1 TO 4 IF button(number) = 2 ^ (8 - number) THEN PRINT “button “; number; “ is open” ELSE PRINT “button “; number; “ is pressed” END IF NEXT number That’s it for this month. If you’re not sure about any of the foregoing, then read the article again – this topic is fairly complex the first time you come across it but it gets easier once you become more familiar with it. Next time, we’ll look at installing a games card into a PC and give some guidelines on using the in-built 5V SC power supply. February 1994  81 VINTAGE RADIO By JOHN HILL Building a simple 1-valve receiver This month, we are going to take a break from our usual format & describe the construction of a simple 1930s-style 1-valve regenerative receiver. It uses a type 30 triode valve & just a handful of other parts. Some time ago (in the November 1991 issue of SILICON CHIP), I wrote about a home-made 2-valve radio receiver aptly named the “Junk Box 2”. It was a simple regenerative set that was built from carefully selected vintage radio parts, thus giving it a reasonably authentic, made-50-years-ago appearance. To achieve this so called authentic look, the parts used in the set’s con­struction were mostly from the mid 1920s to early 1930s – the type of cast-off equipment a young radio enthusiast would have had in his junk box during the 1940s era. The Junk Box 2 story went over fairly well and I personally know of four collectors who were interested enough to build 2-valve regenerative sets of their own. Even at the time of writ­ing, I am still receiving mail regarding the Junk Box 2. However, the general feeling was that Junk Box 2 was unique. Duplicating it was almost impossible for most would-be constructors, due mainly to the lack of appropriate old-style vintage radio parts – vernier dials, The author’s experimental regenerative set has two front panels. The unattached panel has potentiometer controlled reaction while the other has capacitor controlled reaction. 82  Silicon Chip base board valve sockets and audio transformers in particular. Another problem for many was the non-availability of high impedance headphones which are rather scarce these days. Most old headphone sets have open circuit windings and require expensive repairs. With these thoughts in mind, I recently set about designing a similar home-built receiver project that would use more readily available components. It is meant to be a working receiver rather than a replica of something from a bygone age. Headphones To solve the headphone problem, an output transformer has been incorporated into the receiver, thereby allowing low im­ pedance 8Ω stereo headphones to be used instead of the hard-to-get high-impedance types. These modern headphones also give better sound reproduction and are more comfortable than the old style types with their hard bakelite earpieces. Because approximately 80% of Australians live in a capital city environment, a 2-valve set is of little advantage and, in most instances, a single valve receiver is quite adequate. A regenerative receiver lacks the ability to separate powerful local stations from weak distant ones and so the set is mainly intended for big city use where a number of local sta­ tions, of roughly equal strength, are available. These local stations should be strong enough to give good performance on a relatively short indoor aerial. As for the little 1-valver pull­ing in distant signals between the powerful locals – well that is simply asking too much from a simple regenerative receiver, even if an extra valve was to be added. It’s a different situation in my country locality in cen­tral Victoria, with only one local station to contend with. Melbourne, Adelaide, Sydney and even a few Queensland and Tasman­ ian stations can be received on this single valve outfit using a 25-metre long aerial and a good earth. So you can use the set to pull in distant signals, provided that there are not too many local stations. Fig.1: the circuit is based on an old 30 triode valve. The “B” battery voltage can be from 18-45V, while the “A” battery voltage is 2V. ANTENNA 100pF T1 M1100 RFC 100pF 400pF V1 30 250pF 2M PHONES Circuit details The circuit for our 1-valver (see Fig.1) is a time-proven one and it contains no mysteries or modifications apart from the output transformer. It is basic and straightforward and can use just about any battery-operated triode valve. I used an old 30, mainly because there are quite a few in my miscellaneous valve box. The octal equivalent of the 30 is the 1H4 and this should also work OK for this type of receiver. If you want the option of adding a second valve later on, a 1J6 twin triode or a 1D8 triode pentode would allow for expansion and additional experimentation if so desired. From a personal point of view, I find building simple regenerative sets quite a satisfying project and it never fails to amaze me how well they perform, especially when one considers the measly number of components used in their construction. I have built many receivers of this nature and have a spe­cial experimental baseboard and front panel which is used when developing one of these little radios. The front panel houses a tuning capacitor, a reaction capacitor, an on/off switch and a phone jack. An experimental circuit board can be screwed to the base board in a matter of minutes and quickly wired to the con­trol panel components. Assembly of these simple radios is not critical and if the components are poorly placed with jumbled wires running back and forth, it seems to make little difference to the set’s operation. However, a neat, well-planned layout always looks better and is less likely to cause trouble with stray coupling. When building a regenerative receiver, it is normal prac­tice to handwind the coil (actually three separate coils wound on the one former). The cardboard rolls used inside Gladwrap® and Alfoil® make excellent coil former material. Winding the coil is one of the most A+ A- B- B+ Rear view of the front control panel (from left): tuning ca­pacitor, on/off switch, phone jack & reaction capacitor. Masonite is quite suitable for circuit board construction when building simple regenerative receivers. This view shows the predilled board with some of the major components in the back­ground. February 1994  83 Fig.2: here are the winding details for the hand-wound coil. Be prepared to experiment with the number of turns & note that all coils are wound in the same direction. The aerial and earth terminals, the aerial plug & two sockets which are connected to the aerial taps on the coil are all mounted directly on the circuit board. Two 9V batteries plugged into a home-made holder are used to make a compact & convenient B battery. B potentials of up to 45V can be connected to the terminals in the foreground. Higher B voltages will give better performance, provided the reaction winding is correctly adjusted. time-consuming aspects of the exercise. Every coil, as wound by various constructors, will differ due to variations in former diameter, gauge of wire and type of wire insulation. Also, the capacitance of the tuning capacitor, the capacitance of the reaction capacitor, the effec­tiveness of the radio frequency choke (RFC) and the type of valve and the plate voltage used to operate it will all affect the optimum space between the windings and the 84  Silicon Chip number of turns in each winding. For these reasons, one can give only a rough indication of the number of turns required when winding the coil (see Fig.2) and leave it to the individual constructor to alter the specifi­ cations to suit each receiver, with its own particular compon­ents. These must be found by trial and error. By the time the correct number of turns for each of the three windings has been established, the coil can be so untidy and messy (with joins etc.) that it may justify a fresh start on a new former. Problems likely to be encountered with an unsuitable coil are as follows: (1). Tuning too broad or too sharp. To correct this problem, either remove a few turns from the aerial coil to sharpen the tuning or add a few turns to broaden it. Tapping the coil could be an advantage (so that different taps can be selected on an experimental basis). (2). Tuning range not centred on the broadcast band. Add turns to the grid or tuning coil if coverage is incomplete at the low frequency end of the dial (tuning gang closed). Remove a few turns if coverage is incomplete at the high frequency end (tuning gang open). (3). Too much or not enough reaction (regeneration). Remove a few turns from the reaction coil to decrease reaction, or add turns to increase reaction. Altering the valve’s plate (B) supply voltage can also alter the reaction effect. In my prototype, I avoided all the hassles of coil winding by using a commercially made Reinartz coil. I have had this coil from my boyhood days but have only recently rediscovered it. The factory made coil has a number of distinct advantages, so if you have one, use it! The commercial coil is relatively small and is housed in an aluminium can which makes mounting much easier. It has a tapped aerial coil for either long or short aerials and the grid and reaction coils are wound with VINTAGE RADIO We are moving in February 1994 MORE SPACE! MORE STOCK! Radios, Valves, Books, Vintage Parts BOUGHT – SOLD – TRADED Send SSAE For Our Catalogue Rear view of the finished receiver. The commercially made Reinartz coil is compatible with tuning capacitors of various sizes & can be replaced with a hand-wound coil if necessary. Note the four brass terminals for the “A” supply & external “B” supply connections. “Litz” wire. It also has an adjust­able iron slug which can be used to centre the coverage on the broadcast band, according to the tuning capacitor used. The prototype receiver worked reasonably well on an 18V “B” supply and a special battery holder was made and attached to one end of the circuit board, thus keeping the B battery self- con­tained within the set itself. Two terminals were also fitted to the circuit board so that the receiver can be operated at other B voltages. In fact, reception is stronger at 45V but the reaction control is rather touchy and more difficult to operate at these higher voltages. When operating with an external B battery, it is necessary to remove the two 9V batteries from their holders. The plate current is 1.7mA at 45V. A 2V filament (A) supply should be used for a type 30 valve but note that other valve types may require different filament voltages. The filament voltage is derived from an external regu­lated supply and this should be capable of delivering 60mA. Headphone connections A few comments about the head­ phone connections may be in order at this stage. The headphone jack or socket must be a stereo type and not a mono type unless mono headphones are to be used. If a stereo socket is wired correctly it will work (in both earpieces) using either stereo or mono headphones. Connect the transformer secondary to the phone socket so that the socket connects the leads to the tip of the phone plug and to the short insulated section immediately next to it. When wired in this manner, 8Ω stereo headphones become 16Ω mono head­phones. Ordinary low impedance mono phones will also work normal­ly with this socket set up. The output transformer was obtained from Dick Smith Electronics, although similar types are available from other suppliers. It is described as an audio line transformer, Cat No. M1100, and has a 5kΩ primary winding (tapped at 2.5kΩ) and a 16Ω secondary winding tapped at 2Ω, 4Ω and 8Ω. It can be used quite successfully in a valve receiver of this kind. Make sure that the appropriate tap is used. Using 16Ω phones on the 8Ω tap lowers the volume as compared to using the 16Ω tap. We need all the output we can get from a 1-valver so don’t loose any by using the wrong tap. Use the 5kΩ primary winding. (In an emergency, or as a temporary measure, it is worth trying a 5000Ω or 7000Ω speaker transformer from an old valve set. Ed.) The control panel on my set has a WANTED: Valves, Radios, etc. Purchased for CASH RESURRECTION RADIO Call in to our NEW showroom at: 242 Chapel Street (PO Box 2029), Prahran, Vic 3181. Phone: (03) 5104486; Fax (03) 529 5639 Silicon Chip Binders These beautifully-made binders will protect your copies of SILICON CHIP. They are made from a dis­tinctive 2-tone green vinyl & will look great on your bookshelf. Price: $A14.95 (incl. postage in Australia). NZ & PNG orders add $5 each for postage. Not available elsewhere. Send your order to: Silicon Chip Publications PO Box 139 Collaroy, NSW 2097 Or fax (02) 979 6503; or ring (02) 979 5644 & quote your credit card number. February 1994  85 The old 30 type valve from the early 1930s has been a popular choice for single-valve radios, such as the one described in this article. Its filament voltage is rated at 2V while the plate voltage can go as high as 45V. compact single gang tuning capacitor, which is both neat and convenient. It is also almost totally unobtainable today and no electronics shop would stock them. However, many an early transistor radio has useable capacitors for this type of application, even if they are double-gang units. Some transistor radios have tuning capacitors of approximately 400/400pF capacitance, while others have much smaller capacitors of about 250/90pF. These latter types are ideal for use as reaction capacitors. A radio frequency choke is not a problem if you don’t happen to have one. A couple of hundred turns of fine wire around a former about the A leftover from the author’s boyhood days: a commercial Reinartz coil. It avoids the hassles of coil winding & looks much neater. In many cases, however, you will have no choice but to wind your own coil. size of a pencil should do the trick. Failing that, buy one at Dick Smith Electronics when you purchase the M1100 transformer. A 2.5mH type should do the job OK. Dick Smith Electronics also stocks the vernier dial used on the prototype’s control panel (Cat No. P-7170.) Terminals or Fahenstock clips for the battery, aerial and earth connections always make a home made receiver look neater. Wires hanging out the back for battery connections look a bit rough and ready and cause short circuits and other problems. Using the set If unfamiliar with a regenerative Using an output transformer & modern 8Ω stereo headphones solves the problem of obtaining hard-to-come-by high-impedance phones. The modern headphones are far more comfortable than the old bakelite types & give much better sound quality. 86  Silicon Chip receiver, it is necessary to appreciate that the reaction control is not simply a volume control, although it does perform that function. More precisely, it increases the gain and the selectivity by using amplified signals from the plate circuit to overcome the various losses in the grid circuit. The best performance is obtained with the regeneration advanced as far as possible, before oscillation (squealing) occurs. In a set that has been properly set up, the reaction should be arranged (according to the number of turns on the reaction coil) so that the receiver breaks into oscillation when the reaction capacitor is about two-thirds closed. This will cover variations from one end of the dial to the other. More reaction is required at the low frequency end of the dial. Avoid oscillation as much as possible because it will be transmitted to nearby receivers and cause interference. So there it is – Junk Box 1 has been built from less junk and contains more readily accessible parts. If you didn’t build Junk Box 2 because of the parts problem, then this simpler pro­ject may SC appeal to you. Please note that the author is not in a position to supply vintage radio parts or circuit diagrams. Any correspondence requiring a reply should be accompanied by a stamped self-ad­ dressed envelope. PRODUCT SHOWCASE Digital multimeter has dual display & bar graph This 3¾ digital autoranging multimeter from Altronics has a wide array of functions which includes capacitance, frequency and hFE measurements. The large liquid crystal display has two 4-digit readouts, an analog bar graph and a large number of annunciators. Among the huge array of multimeters available on the market, the BX-905A stands out as having a very useful range of measurement functions, 16 in total. With most of these functions, special features can be selected such as automatic hold, a relative mode, and a maximum and minimum recording mode with up to 5 memories. Readings are shown on a 14.5mm high 4-digit main display with a smaller 12mm 4-digit display for dual function readings. Both 4-digit displays can indicate up to 3999. The bar graph works in conjunction with the main display and is useful for indicating changes and trends in variable readings. The graph is graduated from 0 to 4 with 32 individual bars. The smaller digital display indicates the range of the bar display: 4, 40, 400 or 1000. The BX-905A multimeter is housed in a tough yellow-orange plastic case measuring 88.5 x 190 x 27.5mm. It weighs about 330g and is supplied with a set of test leads, a 9V battery (installed) and an operating manual. Front panel controls are the rotary function switch which also doubles as an off switch and seven pushbutton switches for special feature selection. There are four input sockets for the multimeter probes; common, the VW and Diode input and the 20A and mA current inputs. An 8-pin circular socket provides for transistor hFE measurements. DC voltage accuracy is claimed as ±0.5% + 1 digit up to 400V and ±0.5% + 2 digits for the 1000V range. The input impedance is 100MW for the 400mV range and 10MW for all other ranges. Maximum input voltage is 1000VDC or 750VAC. AC voltage accuracy is claimed to be within ±0.8% + 3 digits. No frequency response is given in the manual, however we measured the AC response at -3dB down (70.7%) at 5kHz. Low frequency response is limited by the two-second update time of the meter which sets a minimum frequency for a steady reading at around 20Hz. DC current accuracy is ±1.0% + 2 digits for readings up to 2A and ±2% + 20 digits for the 20A range. AC current accuracy is a little higher at ±1.5% + 5 digits up to 2A and ±2.5% + 20 digits for the 20A range. Maximum input voltage for current measurement is 60V DC or 25V AC. The voltage drop across the meter when measuring current (“burden” voltage) is 400mV. Resistance accuracy is ±0.8% + 2 digits for the 400W range and ±0.5% + 2 digits for the 4kW to 4MW ranges. The 20MW range is ±0.8% + 10 digits. Note that the current applied to the resistor under measurement is low enough so that the resultant voltage is below the 0.6V turn on voltage of silicon diodes and transistors. This means that for most measurements you can test resistors while they are still in-circuit. This is quite handy for servicing work. Frequency measurement accuracy is ±0.5% + 1 digit for 10Hz up to 2MHz. The signal must be greater than 1V RMS for frequencies between 10Hz and 100Hz and more than 500mV RMS for frequencies above 100Hz. Capacitance measurement is from 10nF (0.01µF) up to 99.9µF while accuracy is ±3.0% +10 digits. While this is be useful, the lack of ranges from 1pF to 1000pF does make the capacitance function less than ideal. We should note that this criticism applies to most digital multimeters. February 1994  87 Neat 4-channel microphone mixer One big problem with today’s cassette decks is that they usually do not provide facilities for microphones, or if they do, they only cope with a stereo pair. This is where this neat little mixer from Avico Electronics comes into its own. It can be regarded as a four channel stereo mixer or an eight channel mono mixer. It has eight 6.35mm jack sockets for low impedance microphones and two additional jack sockets for the left The diode test function applies a forward DC current of about 1mA to the device under test. The resulting forward drop is displayed on the meter. The test is also suitable for checking transistor junctions and LEDs although the test current is usually insufficient for a LED to produce any significant light output. This function also provides an audible buzzer which can be selected if required. It sounds whenever the resistance is less than about 30W – good for continuity checking on cables and smilar work. The hFE measures transistor DC gain up to 1000. Both NPN and PNP types can be measured in the test socket which caters for transistors with EBC and BCE pin-out configurations. Larger transistors such as those in TOP-3 and T0-3 packages will need to be connected with hookup-wire to the socket. The BX-905A has many LCD an88  Silicon Chip and right stereo outputs. On the front panel there are ten knobs, eight as the individual microphone level controls and two as the master output controls. It is housed in a sturdy steel case with tiny dimensions: 150mm wide, 110mm deep and 55mm high. It is powered by an internal 9V battery or a 9V DC plugpack. Input impedance is 600W and sensitivity is 5mV in for 90mV out. For further information, contact Avico Electronics Pty Ltd, Unit 4/163 Prospect Highway, Seven Hills NSW 2147. Phone (02) 624 7977. nunciators to indicate such selections as the measurement units and the selected mode. When measuring volts, for example, the units displayed are mV or V, with a minus sign indicating negative voltages. An AC annunciator indicates when the AC ranges are selected. If auto-ranging is in effect, AUTO is displayed. You can also select ranges manually with the range switch and then the AUTO annunciator drops out. The MIN/MAX switch enables recording of maximum and minimum values for the function selected. When this feature is selected the MIN MAX annunciator is displayed. To display the minimum value, press the MIN/ MAX switch and for the maximum value press the MIN/MAX switch again. The annunciator indicates whichever reading is displayed. The RELative switch selects difference measurement between the reading displayed the instant the REL switch was pressed and the current input voltage. A small triangle annunciator is displayed in this mode. In auto hold mode, the current reading is frozen on the display. Press the AUTO H switch and the AH annunciator is displayed along with the last reading. The MEMORY switch enables storage of up to five measured values, as indicated by the MEM annunciator. Pressing the RECALL switch brings up the corresponding annunciator and the stored data is displayed. The SHIFT key enables the different features available on a particular function to be displayed. When the ACV function is selected, the shift key selects dB. The display now indicates in dBm where the result is in dB with respect to 1mW into a 600W load. Other shift key functions activate the diode test, hFE and 20A for ACA and DCA measurements. Normally, the smaller digital display shows the selected range of the meter, but when special features are selected the display can show the input value when in the relative and maximum/ minimum modes and the range when in auto hold mode. Meanwhile, the main display shows the relative value, the min, max or the hold value respectively. A special limit feature has also been added to this multimeter. It allows setting of two values, one a high value and one a low value. When the measured value exceeds the high value, the display shows HI. If the value is below the low value the display shows LO and if the value is between the high and low values the display shows PASS. The annunciator shows LIMIT in this mode and the smaller digital display indicates the current reading. The small manual supplied with the BX-905A multimeter provides all the necessary information and detail to allow the owner to become acquainted with the features of this rather complex multimeter. One specification which was not mentioned in the manual is expected battery life. We measured the current drain at 5.5mA which is rather high and this will correspond to about 70 hours of use with a standard zinc-carbon battery. We would strongly suggest the use of an alkaline battery. The meter does have automatic power-down after about 20 minutes of operation. The BX-905A multimeter is priced at $199. An optional holster is available for $15.95 and a carry case for $12.50. These are available from Altronics, 174 Roe Street, Perth WA 6000. Phone (09) 328 2199. Battery eliminator has 850mA output Panasonic’s snap video camera This streamlined battery eliminator is intended for laptop computers, amateur transceivers, car radios and CD players and other appliances which draw higher currents than can be supplied by most plugpacks. It is switchable between 3, 4.5, 6, 7.5, 9 and 12V DC and can deliver up to 850mA. On good feature about the voltage selection is that the slide switch is on the plug side, so it cannot be changed without pulling it output of the socket. This means that there is no chance of the voltage inadvertently being changed while a device is powered up. Another good feature is the red LED which tells you that the unit is powered. The unit has a solid state regulator and overload protection. It comes with four DC plugs, making it compatible with a wide range of DC devices. It is approved to Australian Standard AS-3108. Designated RBE850, the battery eliminator is priced at $54.95 and is distributed by Avico Electronics Pty Ltd, Unit 4/163 Prospect Highway, Seven Hills NSW 2147. Phone (02) 624 7977. A new concept in camcorders, the Panasonic CS1 is designed to run on AA-size alkaline batteries. When the rechargeable batteries run out, the user can attach the alkaline battery case to the camcorder to obtain an extra 60 minutes of video taping. A combination of rechargeable and alkaline batteries can give you up to 135 minutes of recording time and recording with the CS1 is simple. All you need to do is press one button to start and release it to stop recording. The CS1 also features a super wide angle and telephoto lens settings. Wider than on previous models, the 28mm wide angle lens allows you to easily shoot wide scenes, especially indoors, without the need to pan or tilt the camera. For dramatic close-ups, switch to a 3x telephoto lens. The introduction of a self-timer function means the person filming can also be in the picture. When the self-timer is on, the CS1 pauses for 10 seconds and then shoots for 10 seconds. Other features include an LCD screen that shows battery and tape life as well as an auto date recorder, an optical direct finder plus an anti-scratch body. The CS1 uses compact-VHS video tapes, which can be played back and edited by putting them in the supplied cassette adaptor, which fits into a normal VHS VCR. For further information, see your local Panasonic retailer. Entry level programmer DATA I/O have designed and manufactured a programmer for the design engineer or for small volume programming. Capable of programming EPROMSs, (E)EPROMs, PROMs, PALs, FPGAs and MICROs from DIP to PLCC, SOIC, QFP and TSOP packages, it performs Load, Program, Verify, Sumcheck, ID test, Illegal bit test, Blank check, Erase electrically erasable devices, continuity check and PLD testing to 4ns speed. The Menu driven user interface makes it easy to operate from a PC and it has a built in full screen editor for editing EPROM data in Hex and ASCII format. For a complete list of supported devices, please contact Nilsen Instruments, PO Box 30, Concord, NSW 2137. Phone (02) 736 2888. SC February 1994  89 Silicon Chip Up To Four Separate Messages); UHF Remote Switch; Balanced Input & Output Stages; Data For The LM831 Low Voltage Amplifier IC; Installing A Clock Card In Your Computer; Index to Volume 2. BACK ISSUES September 1988: Hands-Free Speakerphone; Electronic Fish Bite Detector; High Performance AC Millivoltmeter, Pt.2; Build The Vader Voice; Motorola MC34018 Speakerphone IC Data; What Is Negative Feedback, Pt.4. November 1988: 120W PA Amplifier Module (Uses Mosfets); Poor Man’s Plasma Display; Automotive Night Safety Light; Adding A Headset To The Speakerphone; How To Quieten The Fan In Your Computer. December 1988: 120W PA Amplifier (With Balanced Inputs), Pt.1; Diesel Sound Generator; Car Antenna/Demister Adaptor; SSB Adaptor For Shortwave Receivers; Why Diesel Electrics Killed Off Steam; Index to Volume 1. March 1989: LED Message Board, Pt.1; 32-Band Graphic Equaliser, Pt.1; Stereo Compressor For CD Players; Amateur VHF FM Monitor, Pt.2; Signetics NE572 Compandor IC Data; Map Reader For Trip Calculations; Electronics For Everyone – Resistors. April 1989: Auxiliary Brake Light Flasher; Electronics For Everyone: What You Need to Know About Capacitors; Telephone Bell Monitor/ Transmitter; 32-Band Graphic Equaliser, Pt.2; LED Message Board, Pt.2. May 1989: Electronic Pools/Lotto Selector; Build A Synthesised Tom-Tom; Biofeedback Monitor For Your PC; Simple Stub Filter For Suppressing TV January 1990: High Quality Sine/Square Oscillator; Service Tips For Your VCR; Speed­ing Up Your PC; Phone Patch For Radio Amateurs; Active Antenna Kit; Speed Controller For Ceiling Fans; Designing UHF Transmitter Stages. Interference; LED Message Board, Pt.3; All About Electrolytic Cap­acitors. June 1989: Touch-Lamp Dimmer (uses Siemens SLB0586); Passive Loop Antenna For AM Rad­ios; Universal Temperature Controller; Understanding CRO Probes; LED Message Board, Pt.4. July 1989: Exhaust Gas Monitor (Uses TGS812 Gas Sensor); Extension For The Touch-Lamp Dimmer; Experimental Mains Hum Sniffers; Compact Ultrasonic Car Alarm; NSW 86 Class Electric Locomotives. September 1989: 2-Chip Portable AM Stereo Radio (Uses MC13024 and TX7376P) Pt.1; Alarm-Triggered Telephone Dialler; High Or Low Fluid Level Detector; Simple DTMF Encoder; Studio Series 20-Band Stereo Equaliser, Pt.2; Auto-Zero Module for Audio Amplifiers (Uses LMC669). October 1989: Introducing Remote Control; FM Radio Intercom For Motorbikes Pt.1; GaAsFet Preamplifier For Amateur TV; 1Mb Printer Buffer; 2-Chip Portable AM Stereo Radio, Pt.2; Installing A Hard Disc In The PC. November 1989: Radfax Decoder For Your PC (Displays Fax, RTTY & Morse); FM Radio Intercom For Motorbikes, Pt.2; 2-Chip Portable AM Stereo Radio, Pt.3; Floppy Disc Drive Formats & Options; The Pilbara Iron Ore Railways. December 1989: Digital Voice Board (Records February 1990: 16-Channel Mixing Desk; High Quality Audio Oscillator, Pt.2; The Incredible Hot Canaries; Random Wire Antenna Tuner For 6 Metres; Phone Patch For Radio Amateurs, Pt.2; PC Program Calculates Great Circle Bearings. March 1990: 6/12V Charger For Sealed Lead-Acid Batteries; Delay Unit For Automatic Antennas; Workout Timer For Aerobics Classes; 16-Channel Mixing Desk, Pt.2; Using The UC3906 SLA Battery Charger IC. April 1990: Dual Tracking ±50V Power Supply; VOX With Delayed Audio; Relative Field Strength Meter; 16-Channel Mixing Desk, Pt.3; Active CW Filter For Weak Signal Reception; How To Find Vintage Radio Receivers From The 1920s. June 1990: Multi-Sector Home Burglar Alarm; Low-Noise Universal Stereo Preamplifier; Load Protection Switch For Power Supplies; A Speed Alarm For Your Car; Design Factors For Model Aircraft; Fitting A Fax Card To A Computer. July 1990: Digital Sine/Square Generator, Pt.1 (Covers 0-500kHz); Burglar Alarm Keypad & Combination Lock; Simple Electronic Die; Low-Cost Dual Power Supply; Inside A Coal Burning Power Station; Weather Fax Frequencies. August 1990: High Stability UHF Remote Transmitter; Universal Safety Timer For Mains Appliances (9 Minutes); Horace The Electronic Cricket; Digital Sine/Square Wave Generator, Pt.2. September 1990: Music On Hold For Your Tele­ Please send me a back issue for: ❏ April 1989 ❏ May 1989 ❏ October 1989 ❏ November 1989 ❏ March 1990 ❏ April 1990 ❏ September 1990 ❏ October 1990 ❏ February 1991 ❏ March 1991 ❏ July 1991 ❏ August 1991 ❏ December 1991 ❏ January 1992 ❏ May 1992 ❏ June 1992 ❏ October 1992 ❏ January 1993 ❏ May 1993 ❏ June 1993 ❏ October 1993 ❏ November 1993 ❏ March 1994 ❏ ❏ ❏ ❏ ❏ ❏ ❏ ❏ ❏ ❏ ❏ ❏ September 1988 June 1989 December 1989 June 1990 November 1990 April 1991 September 1991 February 1992 July 1992 February 1993 July 1993 December 1993 ❏ ❏ ❏ ❏ ❏ ❏ ❏ ❏ ❏ ❏ ❏ ❏ November 1988 July 1989 January 1990 July 1990 December 1990 May 1991 October 1991 March 1992 August 1992 March 1993 August 1993 January 1994 ❏ ❏ ❏ ❏ ❏ ❏ ❏ ❏ ❏ ❏ ❏ ❏ ➦ Use this handy form to order your back issues December 1988 September 1989 February 1990 August 1990 January 1991 June 1991 November 1991 April 1992 September 1992 April 1993 September 1993 December 1993 Enclosed is my cheque/money order for $­______or please debit my: ❏ Bankcard ❏ Visa Card ❏ Master Card Signature ____________________________ Card expiry date_____ /______ Name _______________________________ Phone No (___) ____________ PLEASE PRINT Street ________________________________________________________ Suburb/town ________________________________ Postcode ___________ 90  Silicon Chip Note: all prices include post & packing Australia (by return mail) ............................. $A7 NZ & PNG (airmail) ...................................... $A7 Overseas (airmail) ...................................... $A10 Detach and mail to: Silicon Chip Publications, PO Box 139, Collaroy, NSW, Australia 2097. Or call (02) 979 5644 & quote your credit card details or fax the details to (02) 979 6503. ✂ Card No. phone; Remote Control Extender For VCRs; Power Supply For Burglar Alarms; Low-Cost 3-Digit Counter Module; Simple Shortwave Converter For The 2-Metre Band. October 1990: Low-Cost Siren For Burglar Alarms; Dimming Controls For The Discolight; Surfsound Simulator; DC Offset For DMMs; The Dangers of Polychlorinated Biphenyls; Using The NE602 In Home-Brew Converter Circuits. November 1990: How To Connect Two TV Sets To One VCR; A Really Snazzy Egg Timer; Low-Cost Model Train Controller; Battery Powered Laser Pointer; 1.5V To 9V DC Converter; Introduction To Digital Electronics; Simple 6-Metre Amateur Transmitter. November 1991: Colour TV Pattern Generator, Pt.1; Battery Charger For Solar Panels; Flashing Alarm Light For Cars; Digital Altimeter For Gliders & Ultralights, Pt.3; Build A Talking Voltmeter For Your PC, Pt.2; Modifying The Windows INI Files. December 1991: TV Transmitter For VCRs With UHF Modulators; Infrared Light Beam Relay; Solid-State Laser Pointer; Colour TV Pattern Generator, Pt.2; Windows 3 & The Dreaded Un­ recov­erable Application Error; Index To Volume 4. January 1992: 4-Channel Guitar Mixer; Adjustable 0-45V 8A Power Supply, Pt.1; Baby Room Monitor/FM Transmitter; Automatic Controller For Car Headlights; Experiments For Your Games Card; Restoring An AWA Radiolette Receiver. December 1990: DC-DC Converter For Car Amplifiers; The Big Escape – A Game Of Skill; Wiper Pulser For Rear Windows; Versatile 4-Digit Combination Lock; 5W Power Amplifier For The 6-Metre Amateur Transmitter; Index To Volume 3. February 1992: Compact Digital Voice Recorder; 50-Watt/Channel Stereo Power Amplifier; 12VDC/240VAC 40-Watt Inverter; Adjustable 0-45V 8A Power Supply, Pt.2; Designing A Speed Controller For Electric Models. January 1991: Fast Charger For Nicad Batteries, Pt.1; Have Fun With The Fruit Machine; Two-Tone Alarm Module; Laser Power Supply; LCD Readout For The Capacitance Meter; How Quartz Crystals Work; The Dangers When Servicing Microwave Ovens. March 1992: TV Transmitter For VHF VCRs; Studio Twin Fifty Stereo Amplifier, Pt.1; Thermostatic Switch For Car Radiator Fans; Telephone Call Timer; Coping With Damaged Computer Direct­ ories; Valve Substitution In Vintage Radios. February 1991: Synthesised Stereo AM Tuner, Pt.1; Three Inverters For Fluorescent Lights; LowCost Sinewave Oscillator; Fast Charger For Nicad Batteries, Pt.2; How To Design Amplifier Output Stages; Tasmania's Hydroelectric Power System. March 1991: Remote Controller For Garage Doors, Pt.1; Transistor Beta Tester Mk.2; Synthesised AM Stereo Tuner, Pt.2; Multi-Purpose I/O Board For PC-Compatibles; Universal Wideband RF Preamplifier For Amateurs & TV. April 1991: Steam Sound Simulator For Model Railroads; Remote Controller For Garage Doors, Pt.2; Simple 12/24V Light Chaser; Synthesised AM Stereo Tuner, Pt.3; A Practical Approach To Amplifier Design, Pt.2. May 1991: 13.5V 25A Power Supply For Transceivers; Stereo Audio Expander; Fluorescent Light Simulator For Model Railways; How To Install Multiple TV Outlets, Pt.1; Setting Screen Colours On Your PC. June 1991: A Corner Reflector Antenna For UHF TV; 4-Channel Lighting Desk, Pt.1; 13.5V 25A Power Supply For Transceivers; Active Filter For CW Reception; Electric Vehicle Transmission Options; Tuning In To Satellite TV, Pt.1. July 1991: Battery Discharge Pacer For Electric Vehicles; Loudspeaker Protector For Stereo Amplifiers; 4-Channel Lighting Desk, Pt.2; How To Install Multiple TV Outlets, Pt.2; Tuning In To Satellite TV, Pt.2; PEP Monitor For Amateur Transceivers. August 1991: Build A Digital Tachometer; Masthead Amplifier For TV & FM; PC Voice Recorder; Tuning In To Satellite TV, Pt.3; Installing Windows On Your PC; Step-By-Step Vintage Radio Repairs. September 1991: Studio 3-55L 3-Way Loudspeaker System; Digital Altimeter For Gliders & Ultralights, Pt.1; Build A Fax/Modem For Your Computer; The Basics Of A/D & D/A Conversion; Windows 3 Swapfiles, Program Groups & Icons. October 1991: Build A Talking Voltmeter For Your PC, Pt.1; SteamSound Simulator Mk.II; Magnetic Field Strength Meter; Digital Altimeter For Gliders & Ultralights, Pt.2; Getting To Know The Windows PIF Editor. April 1992: Infrared Remote Control For Model Railroads; Differential Input Buffer For CROs; Studio Twin Fifty Stereo Amplifier, Pt.2; Understanding Computer Memory; Aligning Vintage Radio Receivers, Pt.1. May 1992: Build A Telephone Intercom; LowCost Electronic Doorbell; Battery Eliminator For Personal Players; Infrared Remote Control For Model Railroads, Pt.2; Aligning Vintage Radio Receivers, Pt.2. June 1992: Multi-Station Headset Intercom, Pt.1; Video Switcher For Camcorders & VCRs; Infrared Remote Control For Model Railroads, Pt.3; 15-Watt 12-240V Inverter; What’s New In Oscilloscopes?; A Look At Hard Disc Drives. July 1992: Build A Nicad Battery Discharger; 8-Station Automatic Sprinkler Timer; Portable 12V SLA Battery Charger; Off-Hook Timer For Tele­phones; Multi-Station Headset Intercom, Pt.2. August 1992: Build An Automatic SLA Battery Charger; Miniature 1.5V To 9V DC Converter; Dummy Load Box For Large Audio Amplifiers; Internal Combustion Engines For Model Aircraft; Troubleshooting Vintage Radio Receivers. September 1992: Multi-Sector Home Burglar Alarm; Heavy-Duty 5A Drill speed Controller (see errata Nov. 1992); General-Purpose 3½-Digit LCD Panel Meter; Track Tester For Model Railroads; Build A Relative Field Strength Meter. October 1992: 2kW 24VDC To 240VAC Sine­wave Inverter; Multi-Sector Home Burglar Alarm, Pt.2; Mini Amplifier For Personal Stereos; Electronically Regulated Lead-Acid Battery Charger. January 1993: Peerless PSK60/2 2-Way Hifi Loudspeakers; Flea-Power AM Radio Transmitter; High Intensity LED Flasher For Bicycles; 2kW 24VDC To 240VAC Sine­wave Inverter, Pt.4; Speed Controller For Electric Models, Pt.3. February 1993: Three Simple Projects For Model Railroads; A Low Fuel Indicator For Cars; Audio Level/VU Meter With LED Readout; Build An Electronic Cockroach; MAL-4 Microcontroller Board, Pt.3; 2kW 24VDC To 240VAC Sine­wave Inverter, Pt.5; Making File Backups With LHA & PKZIP. March 1993: Build A Solar Charger For 12V Batteries; An Alarm-Triggered Security Camera; Low-Cost Audio Mixer for Camcorders; Test Your- self On The Reaction Trainer; A 24-Hour Sidereal Clock For Astronomers. April 1993: Solar-Powered Electric Fence; Build An Audio Power Meter; Three-Function Home Weather Station; 12VDC To 70VDC Step-Up Voltage Converter; Digital Clock With Battery Back-Up; A Look At The Digital Compact Cassette. May 1993: Nicad Cell Discharger; Build The Woofer Stopper; Remote Volume Control For Hifi Systems, Pt.1; Alphanumeric LCD Demonstration Board; Low-Cost Mini Gas Laser; The Micro­soft Windows Sound System. June 1993: Windows-Based Digital Logic Analyser, Pt.1; Build An AM Radio Trainer, Pt.1; Remote Control For The Woofer Stopper; A Digital Voltmeter For Your Car; Remote Volume Control For Hifi Systems, Pt.2; Double Your Disc Space With DOS 6. July 1993: Build a Single Chip Message Recorder; Light Beam Relay Extender; Build An AM Radio Trainer, Pt.2; Windows Based Digital Logic Analyser; Pt.2; Low-Cost Quiz Game Adjudicator; Programming The Motorola 68HC705C8 Micro­ controller – Lesson 1; Antenna Tuners – Why They Are Useful. August 1993: Low-Cost Colour Video Fader; 60LED Brake Light Array; A Microprocessor-Based Sidereal Clock; The Southern Cross Z80-based Computer; A Look At Satellites & Their Orbits; Unmanned Aircraft – Israel Leads The Way; Ghost Busting For TV Sets. September 1993: Automatic Nicad Battery Charger/Discharger; Stereo Preamplifier With IR Remote Control, Pt.1; In-Circuit Transistor Tester; A +5V to ±15V DC Converter; Remote-Controlled Electronic Cockroach; Restoring An Old Valve Tester; Servicing An R/C Transmitter, Pt.1. October 1993: Courtesy Light Switch-Off Timer For Cars; FM Wireless Microphone For Musicians; Stereo Preamplifier With IR Remote Control, Pt.2; Electronic Engine Management, Pt.1; Mini Disc Is Here; Programming The Motorola 68HC705C8 Micro­ controller – Lesson 2; Servicing An R/C Transmitter, Pt.2. November 1993: Jumbo Digital Clock; High Efficiency Inverter For Fluorescent Tubes; Stereo Preamplifier, Pt.3; Build A Siren Sound Generator; Electronic Engine Management, Pt.2; More Experiments For Your Games Card; Preventing Damage To R/C Transmitters & Receivers. December 1993: Remote Controller For Garage Doors; Low-Voltage LED Stroboscope; Low-Cost 25W Amplifier Module; Peripherals For The Southern Cross Computer; Build A 1-Chip Melody Generator; Electronic Engine Management, Pt.3; Index To Volume 6. January 1994: 3A 40V Adjustable Power Supply; Switching Regulator For Solar Panels; Printer Status Indicator; Mini Drill Speed Controller; Stepper Motor Controller; Active Filter Design For Beginners; Electronic Engine Management, Pt.4; Even More Experiments For Your Games Card. PLEASE NOTE: all issues from November 1987 to August 1988, plus October 1988, January, February, March & August 1989, May 1990, and November and December 1992 are now sold out. All other issues are presently in stock, although stocks are low for some older issues. For readers wanting articles from sold-out issues, we can supply photostat copies (or tearsheets) at $7.00 per article (incl. p&p). When supplying photostat articles or back copies, we automatically supply any relevant notes & errata at no extra charge. February 1994  91 ASK SILICON CHIP Got a technical problem? Can’t understand a piece of jargon or some technical principle? Drop us a line and we’ll answer your question. Write to: Ask Silicon Chip, PO Box 139, Collaroy Beach, NSW 2097. Adding seconds to the Jumbo Clock Regarding your excellent Jumbo Digital Clock published in the November 1993 issue, well, I have only one thing to say – I want seconds! For at least two years I have been searching for a suitable circuit to power six large 7-segment displays I purchased in Singapore, and only now has such a circuit been published. Can a “patch” be added to display seconds? And can your circuit drive the displays I own? Or would modifications have to be made? I have enclosed the specifications of my dis­plays for you to compare. (M. B., Wheelers Hill, Vic). • After checking the specs of your 7-segment displays, we cannot see any reason why they cannot be used. You will however need to change the layout of the board for the displays because of the differing pinouts. To add seconds display to the circuit, you will have to duplicate the circuitry of IC5 and IC6 (the minutes circuitry) and delete the circuitry of IC3, IC4a and diode D4. The 1Hz clock signal from IC2a would then drive the clock input of the units-seconds Will a CD player play CD-ROMs? I wondered if it were at all possible to use a conventional audio CD player (with some modification) with a PC to use CD-ROM discs. What major differences are there between the two types of drive to justify the dramatic difference in pricing? I assume there would need to be more accurate control over the tracking to ensure good error free data transfer. Companies like SEGA seem to make a complete CD-based games machine at a very economical price but bare PC CD-ROM drives seem to be a little overpriced 92  Silicon Chip counter, which in turn would drive the tens-seconds counter. The carry-out pin (pin 5) of this counter would then feed the clock input (pin 1) of IC5 and would then continue on as normal. Because your displays have more LEDs per segment, you may have to experiment with the emitter current-limiting resistors for transistors Q7-27 to obtain a suitable level of brightness. 270Ω would be a good place to start. Are NTSC to PAL converters available? Is there such a thing as a NTSC to PAL converter. If so, is it buildable in kit form, etc? (M. M., Wapparaburra Haven, Qld). • A few years ago, the answer to this question would be that the only standards converters available were those used in TV stations. Now there are simple “analog” converters which allow an NTSC picture to be viewed on a PAL monitor although with reduced picture height. Such a device is available from Av-Comm Pty Ltd for $155 (Cat. No. T-1200). Alternatively, there are digital con­ verters with a “field to me. Would this be a feasible project for SILICON CHIP? (D. S., Alfords Point, NSW) • There are two major differences between CD-ROM drives and CD players. The first, as you suggest, gives better control over tracking but, secondly, there is much better error correction in the CD-ROM drives, to provide error free data. Having said that, it seems that the only real reason for a difference in price is economies of scale – far more CD players than CD-ROM drives are presently being produced. Unfortunately, we don’t think it will ever be a practical project for SILICON CHIP magazine. store” to hold and manipulate the video lines for a complete field. Av-Comm have one of these too, at $950 (Cat. No. T-1400). You can contact Av-Comm by phoning (02) 949 7417. Motors for electric vehicles In the May 1991 issue of SILICON CHIP there was an article on electric and solar powered cars entitled “Motors for Electric Vehicles”. This article mentioned a solar powered car (Solar Star II) which the designer was thinking of putting into production. What I would like to know is if the car ever did go into produc­tion and, if so, where it would be sold and how much it would be sold for. Perhaps you could do a follow up article on this subject, including any other electric/solar cars that are available to the public. (R. S., Old Guildford, NSW). • As far as we know, the solar car mentioned in May 1991 issue did not go into production. Nor are any electric/solar cars presently available in this country on a normal production basis. Wireless microphone queries I am writing to you about your article in the October 1993 issue on the FM wireless microphone. I bought it in kit form from Oatley Electronics and I note that they suggest a 22kΩ resistor instead of the 560Ω bias resis­tor for the electret and to replace this with a 47kΩ trimpot if using the unit as a line transmitter. Even though the 560Ω resis­tor works far better, I would be interested to know why the vast difference, as I am new to electronics and just becoming an addict to it. Also, the pinout diagram for transistor Q1 (BC549) is shown in your article as being viewed from below and with the leads in a triangle configuration. The transistor I received with the kit had the leads in-line and actually, in the end, was put in on the circuit board side so as it would work. According to the archer semiconductor reference guide, the in-line transistors go C, B, E; ie, symbol viewed from bottom. Whose mistake is this? After soldering the components in I had great difficulty removing this transistor to swap the emitter and collector over. The number of the transistor I received seemed to read C549PH87. Is this correct or is there a better transistor avail­able that fits the circuit board’s component positions? Or does it sound like this one could be faulty? Also could I increase signal strength without too much alteration to the signal by adding another audio transistor? (J. C., Cooma, NSW). • The reason for changing the input resistor is twofold. First, the 560Ω resistor (or a similar value ranging up to 2.2kΩ) is necessary to provide drain current for the FET source follower inside the electret microphone housing. The FET is there to buffer the capacitive source impedance of the electret capsule. Second, the size of the resistor sets the gain and for sources which have large signal outputs, the resistor should be increased or changed in favour of a potentiometer at the input. The transistor base diagram we published is correct. The transistor should be mounted so that the flat of its case matches the outline on the screen printed overlay of the PC board. A C549 transistor is the same as a BC549. Revenge of the woofer stoppers I have assembled the Woofer Stopper from a kit (after an agonising 10-week wait for them to ship) and it is working nor­mally. Normally, but not forcefully! At least not forcefully enough for the dog across the street! He can certainly hear it but that’s about all. When I hit the start button I expect the dog to get turned inside-out by the blast! I’m talking revenge here! Well, you know what I mean. Let me pose several questions in a quest to make the device into something effective. Let’s start with the enemy. What is the frequency range of their hearing and at what frequency is their hearing most sensi­tive? Does this frequency Reluctor ignition not effective at low RPM I am enquiring about the reluctor ignition kit featured in your May 1990 issue. I have an example of the kit fitted to my Subaroo EA81 1800cc direct-driven gyrocopter. The propeller is bolted directly to the back of the motor and not through a reduc­tion drive. On the bench the ignition system worked perfectly but on the gyro it wouldn’t fire. All wiring was checked and proved OK. The distributor was removed and hand spun. All worked perfectly until we slowed the spin speed. At about 30-40 RPM (direct on the distributor shaft) the system fired erratically. At speeds less than this it wouldn’t fire at all. This is the crux of the problem, as we start the motor by hand throwing the propeller. Putting in a points distributor isn’t a preferred option, as several gyro pilots I know have tried both points and reluctor pickups (with the original Subaroo black box) and have measured at least 100 RPM increase in overall revs with the reluctor pickup and so increased their flying safety margin. vary from dog to dog? Would it be more effective to sweep a range of frequencies? The DSE catalog (1992) claims the piezo tweeter can handle 40 watts continuous and its response is 5-27kHz. Why then are we delivering only (I’m guessing here) about 2 or 3 watts to the tweeter? How do I go about delivering the full 40 watts? As mentioned above, when I hit the button I want things to happen! I look forward to your response – and I hope that a “brute-force” (pun) version is just around the corner. By the way, my son (13) can clearly tell when the device is operating – even when he is standing in the offending dog’s driveway across the street. (R. N., Auckland, NZ). • We know exactly what you mean when you start talking about revenge. That is what it is all about. Yes you can drive the piezo tweeter harder but The propeller limits my engine revs to approximately 3500 RPM. It has been cut to do so, because at 3548 RPM the 54-inch diameter propeller goes supersonic at the tips, resulting in a loss of thrust (and an incredibly loud and terrible noise). So the prop has been cut so that at full throttle the maximum the motor does is 3500 RPM. What do I have to do to the input circuitry so that the kit will operate at very low shaft speeds but still function at normal speeds up to 3500 RPM? A switched start circuit isn’t a preferred option (but I am willing to build one if necessary) as it would be possible to accidentally leave it in start mode when flying and so possibly cause the trigger IC to fail. (A. W., Tailem Bend, SA). • There is no simple way to modify the circuit to improve the gain and thereby the low-speed response from the reluctor. Howev­er, there are three things you can do to the reluctor itself to improve its response. The first of these is to reduce the gap between the toothed ring and the reluctor coil. The second is to increase the number of turns on the coil itself, and the third is to use a stronger magnet. that may not be enough, so let’s answer your questions in detail. You state that the tweeter is rated at 40 watts and has a frequency response of 5kHz to 27kHz. The problem is that the frequency response of piezo tweeters is never smooth – it is usually very lumpy and unless you have a precise means of measur­ing the actual sound output (difficult above 20kHz), then it is best to stick to the frequency we have chosen which is above audibility for most of the population. By the way, if your son can hear it from across the street, he would find it painful close up, say within several metres. Dogs have a frequency range (supposedly) well up to 40kHz or more but they all vary, just like humans, and many older dogs are deaf. Again, this is an argument for keeping the frequency fed to the tweeter as low as possible. Another reason to keep it low is that piezoelectric tweeters are February 1994  93 Remote control extender is weak I have built the “Remote Control Extender For VCR’s” as described in your September 1990 edition and I have had trouble with it. LED 1 acknowledges that it is receiving the signal sent out by the remote control but it does not retransmit this signal with enough amplitude for the video to pick it up. My first question is which way should the 7808 3-terminal 8V regulator be placed in the circuit? The diagram is different to the picture on page 26? Which is the correct way? What should the voltage and current be across the infrared LED (IRLED 1) and also at various other points so that it is possible to check a capacitive load, typi­cally from .05µF to 0.33µF or more. This means that if you double the frequency fed to the tweeter, you double the current that needs to be delivered by the driving amplifier. The power rating assigned to piezo tweeters is fairly arbitrary. If it is rated at, say, 40 watts, for a given frequency range, that means that it will handle the output of a 40 watt amplifier intended for 8Ω loudspeakers. It does not mean that it can handle 40 watts itself. What it probably means is that it can handle a signal of around 18V RMS on a sinew­ave or around 22V peak. We guess that the ultimate limita­tion on power handling in a piezo tweeter would be mechanical – at some stage the piezoelectric forces will become so high that the ceramic diaphragm will be fractured or damaged in some other way. We think that this means you could probably boost the supply voltage to the woofer stopper to around 22V DC; ie, +V1 on the circuit can be connected to +22V. However the peak current through the Mosfets starts to become very high and it would be wise to connect a 2Ω resistor in series with the DC supply to limit the peak current to a safe value. The effect of this modification will be to increase the power output by about four times. At about +6dB, this is not a big increase and not in keep94  Silicon Chip the rest of the circuit? Since the project was published have there been any chang­es? (A. L., Kenmore, Qld). • The 7808 regulator is shown correctly. The metal tab of the regulator should face away from IC1. As suggested in subsequent Notes & Errata, it may be worthwhile modifying the AGC character­istic of the SL486 (IC1) by changing the 0.15µF capacitor at pin 8 of IC1 to a larger value, say 10µF or up to 47µF. A 22kΩ resis­tor should also connected across it. The current through IRLED 1 depends on the transmit­ter code from your remote handpiece. It can be seen as a high frequency pulse waveform if you connect an oscilloscope at the collector of Q1. ing with the revenge you are seeking. If you want a lot more output, you should go for more tweeters connected in parallel or purchase the most rugged and efficient tweeter you can get. A look at the current DSE catalog shows their model KSN-1177 (Cat C-2204) twin drive tweeter has an efficiency of 99dB and that would be the one to go for since its sensitivity is 6dB more than any other model. Its current Austra­lian price is $39.95. If you use this tweeter and do the mods we suggest above, you should get an effective increase in actual power output of around 12dB or so and this is very worthwhile. Incidentally, don’t try pushing the Woofer Stopper any harder by increasing the supply voltage above 22V. To do so is inviting Mosfet failure. A higher-power circuit would require a completely different drive arrangement. Beware reverse polarity diodes I am building the dual tracking 50V power supply described in the April 1990 issue of SILICON CHIP. After destroying a number of components we have found the polarity on our TO-220 style BY229 to be the reverse of that shown in your article. After turning them around the circuit is performing correctly. Is this an error in the article or is there variation between manufacturers as to the package pinouts. If, as I sus­ pect, it is the latter, what in general do I look for in future to identify correct orientation of components? (P. L., Osborne Park, WA). • We suspect you have been supplied with reverse polarity diodes which would be marked BY229 600R or 800R. The “R” signi­fies reverse polarity. Fast clock for model railroads I would like a digital fast clock to which slave units can be added for use in different locations around my model railway layout, as one clock cannot be seen from all areas of the layout. I have just seen this product review on a fast clock in the September 1993 issue of Model Railroader. It looks like just what we need, only not so many different speed selections. Six times fast is about right and we don’t want all the fancy bits saying “hello” etc. Can you come up with something? (W. H., Glen Innes, NSW). • Have you had a look at the Jumbo Clock circuit published in the November 1993 issue? It has the same circuit as the Classic Clock published in April 1993 except for having large displays and a different board. Either of these clocks could be modified to provide “6x” fast clock operation. A simple modification is required. Disconnect the anode of diode D1 from its present position and connect it to pin 6 of IC3. You could also get a “2x” fast clock but that would be a different modification. You could build up slave clocks to operate from a master but there would be little saving in cost – you might just as well build up a bunch of these clocks and then synchronise them all at the start of an operating session. Jaycar have the Jumbo Clock available in kit form at $109. Jaycar also have had a limited quantity of 4-digit clock modules at the princely price of $2.50. While we do not have circuit details, it should be possible to make them run fast too, even though you may have to add an external timing circuit; eg, using a 555 IC. (Editor’s note: railway modellers frequently use a six times fast clock so that they can simulate a 24-hour day SC in a 4-hour operat­ing session.) MARKET CENTRE Cash in your surplus gear. Advertise it here in Silicon Chip. ANTIQUE RADIO CLASSIFIED ADVERTISING RATES Advertising rates for this page: Classified ads: $10.00 for up to 12 words plus 50 cents for each additional word. Display ads (casual rate): $20 per column centimetre (Max. 10cm). Closing date: five weeks prior to month of sale. To run your classified ad, print it clearly in the space below or on a separate sheet of paper, fill out the form & send it with your cheque or credit card details to: Silicon Chip Classifieds, PO Box 139, Collaroy, NSW 2097. Or fax the details to (02) 979 6503. _____________ _____________ _____________ _____________ _____________ _____________ _____________ _____________ _____________ _____________ _____________ _____________ _____________ _____________ _____________ ANTIQUE RADIO RESTORATIONS: specialist restoration service provided for vintage radios, test equipment & sales. Service includes chassis rewiring, recon­ densering, valve testing & mechanical re­­furbishment. Rejuvenation of wooden, bakelite & metal cabinets. Plenty of parts – require details for mail order. About 1200 radios within 16,000 square feet. Two-year warranty on full restoration. Open on Saturday 10am-4.30pm; Sunday 12.30-4.30pm. 109 Cann St, Bass Hill, NSW 2197 Phone (02) 645 3173 BH or (02) 726 1613 AH. FOR SALE _____________ _____________ _____________ _____________ _____________ _____________ _____________ _____________ _____________ _____________ _____________ _____________ _____________ _____________ _____________ _____________ _____________ _____________ _____________ _____________ _____________ _____________ _____________ _____________ _____________ _____________ _____________ _____________ _____________ _____________ _____________ _____________ _____________ _____________ _____________ THE HOMEBUILT DYNAMO: (plans) brushless, 1000 DC watt at 740 revs. $A85 postpaid airmail from Al Forbes, PO Box 3919 - SC, Auckland, NZ. Phone Auckland (09) 818 8967 any time. Rotor magnets (3700 gauss) kit now available. WEATHER FAX programs for IBM XT/ATs *** “RADFAX2” $35 is a high resolution, shortwave weather fax, Morse & RTTY receiving program. Suitable for CGA, EGA, VGA and Hercules cards (state which). Needs SSB HF radio & Radfax decoder. *** “SATFAX” $45 is a NOAA, Meteor & GMS weather satellite picture receiving program. Needs EGA or VGA plus “WEATHER FAX” PC card. *** “MAXISAT” $75 is similar to SATFAX but needs 2Mb expanded memory (EMS 3.6 or 4.0) ❏ Bankcard   ❏ Visa Card   ❏ Master Card Card No. ✂ Enclosed is my cheque/money order for $­__________ or please debit my RCS RADIO PTY LTD Signature­­­­­­­­­­­­__________________________ Card expiry date______/______ Name ______________________________________________________ Street ______________________________________________________ Suburb/town ___________________________ Postcode______________ RCS Radio Pty Ltd is the only company that manufactures and sells every PC board and front panel published in SILICON CHIP, ETI and EA. RCS Radio Pty Ltd, 651 Forest Rd, Bexley 2207. Phone (02) 587 3491 February 1994  95 TRANSFORMER REWINDS ALL TYPES OF TRANSFORMER REWINDS TRANSFORMER REWINDS Reply Paid No.2, PO Box 438, Singleton, NSW 2330. Ph: (065) 76 1291. Fax: (065) 76 1003. and 1024 x 768 SVGA card. All programs are on 5.25-inch or 3.5-inch disks (state which) & include documentation. Add $3 postage. Only from M. Delahunty, 42 Villiers St, New Farm, Qld 4005. Phone (07) 358 2785. PAY TV & SATELLITE Scrambling News Monthly, with the latest on de­scrambling techniques & addresses, where to buy the latest descramblers. Send stamp for info. John Papp, Box 37885 Winnellie, NT 0821. PRINTED CIRCUIT BOARDS for the hobbyist. For service & enquiries contact: T. A. Mowles (08) 326 5590. SUBSTITUTE FOR A HANDFUL OF ICs: Parallax “BASIC STAMP”. A gen­er­al purpose small circuit module, it is really a 25 x 50mm board with a computer chip (4MHz PIC 16C56), EEPROM, 8 I/O pins, board space includes prototyping area. Program it on a PC (only 33 instructions) with development kit which includes one “BASIC STAMP” ($249 plus S/T & post), extra modules ($66 plus S/T & post). Send 45c stamp for more information. Parallax distributor and techni­cal support in Australia: MicroZed Computers, PO Box 634, Armi­dale, NSW 2350. Facsimile (067) 72 8987. MICASOFT Electronics and Computing tutor program, written in UK, ideal for TAFE, schools, or individual use. Now available in Australia. Send $1.80 in stamps for demo disk (tell us what size). MicroZed Computers, PO Box 634, Armidale 2350. 68705 MICRO EMULATOR!!!: Yes! A fair dinkum 68705 hardware ICE for $285 (B&T $330). Run programs in RAM, builtin disassembler, single step, break points, the works! It even emulates 2716, 2732 and 2764 EPROMs. Can be used with 96  Silicon Chip MEMORY & DRIVES PRICES AT DECEMBER 1ST, 1993 SIMM 1Mb x 3 70ns 1Mb x 9 70ns 4Mb (72-pin) 4Mb x 9 70ns 4Mb x 8 80ns $63 $68 $265 $235 $210 DRAM DIP 1 x 1Mb 256 x 4 1Mb x 4 70ns 70ns Z $8 $8 $35 IBM PS.2 50/55/70 70/35 90/95 2Mb 4Mb 4Mb $136 $265 $265 MAC 2Mb SI & LC 4Mb P’Book $135 $320 CO-PROCESSORS 387SX to 25 387DX to 33 $105 $105 LASER PRINTER HP with 4Mb $260 TOSHIBA T3200SX T44/6400 T5200 4Mb 4Mb 8Mb $360 $340 $680 SUN SPARC 10/20 16Mb $920 DRIVES SEAG 130Mb 16ms $290 SEAG 452Mb 12ms $720 SEAG 1.05GB 10ms $1660 Sales tax 21%. Overnight delivery. Credit cards welcome. Ring for Latest Prices 1st Floor, 100 Yarrara Rd, PO Box 382, Pennant Hills, 2120. Tel: (02) 980 6988 Fax: (02) 980 6991 PELHAM ICL 286 Board Kits All in one board with two serial, printer, IBM keyboard, high density floppy & IDE mono video interface. Up to 4Mb RAM, 80286-16cpu, MS-DOS compatible, 130 page manual, small size 170mm x 255mm. Max I/O kit for PCs, 7 relays, ADC, DAC, stepper driver, TTL inputs, with software $169 PC I/O card with 8255 chip 24 I/O lines programmable as inputs or outputs $69 1.5 watt AM broadcast transmitter XTAL locked $49 2.5 watt FM broadcast transmitter 88-108MHz. $49 Digi-125 audio power amp (over 19,000 sold since 1987) 50 watt/8 $14 125 watt/4 $19 New 200 watt/2 version $29 Infrared relay kit $9 Remote control tester $4 $299 Ampo little PC All in one NEC V40 CPU board, MS-DOS compatible, high density floppy. SCSI hard disk, 2 serial, printer, solid state hard disk, IBM keyboard interface, (4W), CMOS single +5V rail, up to 768Kb RAM, 384Kb ROM, 145mm x 250mm, 98page manual. $299 P.C. Computers 36 Regent St, Kensington, SA. Phone (08) 332 6513. a PC, MAC etc. Optional 687053/U/R ($115) and C4/C8 ($95) programmers for direct connec­tion to 68705 emulator. Kits and further information from Graham Blowes, Mantis Micro Products, 38 Garnet St, Niddrie 3042. Phone (03) 337 1917(ah), (03) 575 3349(bh), fax (03) 575 3369. A TRUE AUSSIE Z80 Development System driven from MS-DOS LPT1. EPROM is emulated during development. PCB and disk full of Source Code, Z8T XASM, Z8TBasic and full circuits. $38. With EPROM $52. Promo disk $2. Don McKenzie, 29 Ellesmere Crescent, Tullamarine 3043. Phone (03) 338 6286. UNUSUAL BOOKS: Electronic Devices, Fireworks, Locksmithing, Radar Invisibility, Surveillance, Self-Protection, Unusual Chem­istry and more. For a complete catalog, send 95 cents in stamps to Vector Press, Dept S, PO Box 434, Brighton, SA 5048. Advertising Index All Electronic Components............8 Altronics ................................ 48-49 Antique Radio Restorations.........95 A-One Electronics........................55 Av-Comm.....................................43 Contan Audio...............................11 David Reid Electronics ..............69 Dick Smith Electronics........... 12-15 Electronic Fault Info.....................19 Emtronics.....................................45 Harbuch Electronics....................69 Instant PCBs................................96 Jaycar ........................ 33-36, 61-64 Macservice....................................3 National Instruments...................25 PC Computers.............................96 Pelham........................................96 Peter C. Lacey Services..............50 Philips Test & Measurement......IBC RCS Radio ..................................95 Resurrection Radio......................85 Rockby Electronics......................60 Rod Irving Electronics .......... 74-78 Silicon Chip Back Issues....... 90-91 Silicon Chip Binders..............80, 85 Silicon Chip Book Club..................9 Tektronix..................................OBC Transformer Rewinds...................96 Wombat Electronics.....................11 Yokogawa..................................IFC FLUORESCENT INVERTER KIT (SC Feb 91) 12V or 24V/5W-21W. 48V version on request. Secondary wind, board plus components $30.00 plus p&p $4.00. Fluorescent inverter kit (SC Nov 93) 12V/24V/48V, 18W and 38W P.O.A. Solar battery charging regulator, short form kit, 12V or 24V (series) (SC Jan 94), employs Mosfet to switch solar array, max current 10A $54.00 plus p&p $4.00. Additional Mosfet $8.00 and Schottky diode $5.00 to make 20A regulator. Cheques and postal money orders accepted with mail orders. Send orders to Otakar Priboj, PO Box 362, Villawood, NSW 2163, Australia. Phone (02) 724 3801 (Otto).